Get documents about "EURO-CV Building new indicators for researchers' careers and
Document Sample
EURO-CV
Building new indicators for researchers’
careers and mobility based on electronic
Curriculum Vitae
FINAL REPORT
JULY 2009
CONTENTS
1. Project Report Summary .............................................................................................. 6
2. Development of electronic CV information systems. Analysis of their potential for
researchers careers and mobility studies ........................................................................ 11
2.1. NORWAY ........................................................................................................... 11
2.1.1. Assessment of databases (work package 1).................................................. 11
2.1.2. Work package 2 (exploratory studies): Data on scientific stays abroad....... 17
2.1.3. Conclusions and recommendations .............................................................. 19
2.2. SPAIN.................................................................................................................. 20
2.2.1. Assessment of databases (work package 1): Introduction to the eCV system
in Spain................................................................................................................... 20
2.2.3. The scientific information system in Andalusia: SICA................................ 25
2.2.4. The CVN-XML standard: Towards a normalized electronic CV................. 26
2.2.5. Exploratory studies (work package 2) .......................................................... 27
2.2.6. Summary of main preliminary findings: ...................................................... 38
2.2.6. Information sources and acknowledgements................................................ 39
2.3. PORTUGAL........................................................................................................ 40
2.3.1. Assessment of existing CV databases (Work Package 1) ............................ 40
2.3.2. The DeGóis Platform.................................................................................... 42
2.3.3. Exploratory studies (Work Package 2)......................................................... 47
2.3.4 Preliminary empirical analysis ...................................................................... 52
Appendix: Characterisation of generic sample....................................................... 57
3. Ad-hoc CV databases: the UK experience ................................................................. 61
3.1. Availability, characteristics and accessibility of electronic CV databases and
registers....................................................................................................................... 61
3.1.1. Research funding organisation and requirement processes.......................... 62
3.1.2. Storage format and policy purposes ............................................................. 65
3.1.3. Uploading culture and level of standardisation ............................................ 65
3.2. Methodological aspects of CV collection, coding and analysis in the UK ......... 65
3.2.1. Collection methodology ............................................................................... 66
3.2.2. Other sources used to complete CV information.......................................... 67
3.2.3. Coding methodology .................................................................................... 68
3.2.4. Characteristics of the sample........................................................................ 71
3.2.5. Analysis ........................................................................................................ 71
4. CV database mapping in other European and associated countries ........................... 75
4.1. FRANCE ............................................................................................................. 75
4.2. ISRAEL ............................................................................................................... 76
4.2.1. Introduction .................................................................................................. 76
4.2.2. National Level .............................................................................................. 77
4.2.3. Institutional Level CVs collection in the Research Universities.................. 77
4.2.4. Researchers CVs Format .............................................................................. 79
4.2.5. SNI’S Selected Registers for the Prime Euro-CV Project............................ 80
4.2.6. CV Coding Issues ......................................................................................... 82
4.2.7. Policy recommendations .............................................................................. 82
4.3. SWITZERLAND................................................................................................. 83
4.3.1. Introduction .................................................................................................. 83
4.3.2. General situation........................................................................................... 84
4.3.3. Universities without central databases containing CV-information............. 86
4.3.4. Universities with a central database containing CV-information................. 89
2
4.3.5. Available information and possible analysis ................................................ 91
4.3.6. Conclusions: Possible future developments and recommendations ............. 95
5. List of Project Participants ......................................................................................... 97
6. Project outcome and diffusion activities: publications and presentations.................. 98
ANNEX 1: Project Seminar Agendas .......................................................................... 100
ANNEX 2: Database Templates................................................................................... 104
3
LIST OF TABLES
Table 1: Institutions using CV electronic information systems in Spain ....................... 22
Table 2: CV electronic information systems in Spain.................................................... 23
Table 3: Distribution of the sample according to discipline........................................... 29
Table 4: destinatios by decade........................................................................................ 33
Table 5: Summary by discipline..................................................................................... 35
Table 6: Fields in FCT-SIG Curriculum Vitae............................................................... 41
Table 7: Production indicators in the DeGóis platform.................................................. 45
Table 8: DeGois data requested and supplied ................................................................ 49
Table 9: Dimensions of Mobility ................................................................................... 52
Table 10: Variables from Plataforma DeGóis used in the analysis................................ 54
Table 11: Variables from Plataforma deGois for the experimental analysis of trajectories
........................................................................................................................................ 55
Table 12: Main research funding organizations ............................................................. 64
Table 13: Researchers' mobility variables...................................................................... 70
Table 14: Distribution of the researchers by discipline and response rate of e-mails that
reached a valid address ................................................................................................... 71
Table 15: ......................................................................................................................... 78
Table 16: Sources for coding CVs.................................................................................. 80
Table 17: Coded information.......................................................................................... 80
Table 18: Curricular information available in the databases of the University of Lucerne
and the University of Neuchâtel ..................................................................................... 91
4
LIST OF FIGURES
Figure 1: Frequency of visits abroad per person, scientific staff only. .......................... 17
Figure 2: Frequency of destinations – all visits.............................................................. 18
Figure 3: Frequency of scientific visits abroad in 2005, 2006 and 2007 ....................... 18
Figure 4: Length of stay abroad...................................................................................... 19
Figure 5: The CVN information system structure .......................................................... 27
Figure 6: Distribution of visits in terms of length .......................................................... 30
Figure 7: Distribution of visits per researcher ................................................................ 31
Figure 8: Countries more often visited ........................................................................... 32
Figure 9: Visits per researcher and year ......................................................................... 33
Figure 10: Countries with significant changes amongst decades ................................... 34
Figure 11: DeGóis data categories.................................................................................. 49
Figure 12: Institutional bias: Universities and other research organisations in the sample
(N=500) .......................................................................................................................... 57
Figure 13: Distribution by scientific fields (N=825)...................................................... 58
Figure 14: Age distribution of sample (N=493) ............................................................. 58
Figure 15: Year completed PhD (N=493) ...................................................................... 59
Figure 16: International mobility for the PhD: main countries ...................................... 59
Figure 17: International mobility for the PhD: scientific fields (N=226) ...................... 60
Figure 18: Date started current position (N=500) .......................................................... 60
Figure 19: Distribution by career (N=500)..................................................................... 60
Figure 20: Career position (N=496) ............................................................................... 61
Figure 21: Collection methodology................................................................................ 68
Figure 22: Graphs 1 to 4 ................................................................................................. 74
Figure 23: Geographical mobility of professors in Communication sciences in
Switzerland: place of doctorate place of employment............................................... 94
5
1. Project Report Summary
The main objective of the EURO-CV Project was to collectively explore the
possibilities of using researchers’ curriculum vitae (CVs) in electronic form (e-CV) as a
way to cope with the lack of data concerning mobility and professional career
trajectories of European researchers. The project was therefore mostly methodologically
oriented, with the specific goals of: (i) taking stock of efforts to build CV databases in
the participating countries; and (ii) collecting electronic CV data in order to conduct
some initial exploratory studies on the possibilities of building mobility indicators.
The project activities were developed between January 2008 and June 2009, with the
participation of seven countries: France, Israel, Norway, Spain (coord.), Switzerland,
Portugal and the United Kingdom. In addition the project team collaborated with
international colleagues with CV analysis or database development experience in
Australia, Argentina and the United States.
All project partners shared a common interest in the EURO-CV goals, which were
initially designed homogenously for all participating countries. The development of the
project and the database mapping exercise that was first conducted showed, however,
that the differences in the national situations concerning the collection and storage of
electronic researcher CVs were substantial. We may classify the national situations of
the participating countries in three major groups: (i) countries in which an electronic CV
information system with national coverage is under development – Norway, Spain and
Portugal; (ii) a country in which no coordinated national effort to collect e-CVs has
been undertaken, but in which an ad-hoc CV database developed by a research team
was available to conduct some exploratory studies – the UK; and (iii) countries in which
no coordinated national effort to collect e-CVs has been undertaken and in which no
large enough ad-hoc database was available – Israel, France and Switzerland. The
contribution of each partner to the project’s results was therefore framed by their
specific national circumstances.
Teams of countries in which an e-CV national information system is under development
focused their contribution on describing the characteristics of the systems and their
potentialities to build mobility and career development indicators through the analysis
of data sets obtained from the corresponding institutions.
National e-CV systems
Norway
Norway is currently building up a national system of information about public research
called the Norwegian Science Index, with the purpose of developing a knowledge base
about Norwegian public research activities in the higher education system. It is likely
that, in the future, research institutes will also be integrated into the system. The plan
for the Science Index included a CV module and a first version was launched in Spring
2009. This initiative will contribute to reducing the fragmentation of information that
6
followed the emergence of uncoordinated local initiatives in universities and state
colleges. A major driving force for constructing an integrated system has been the
gradual introduction of ‘result based core funding’ in higher education. The National
Science Index system will mainly result from merging two major pre-existing
information systems: BIBSYS and FRIDA. Only the latter continued to develop its
initial e-CV module over the years. Therefore the EURO-CV Norwegian report focuses
on the characteristics and data from this system, which currently covers 75% of higher
education institutions and will, very likely, cover 100% in the near future. Data was
requested from FRIDA on career trajectories and international visits of researchers. The
sample obtained included 12,000 researchers (after disregarding guests and emeritus
professors). An important methodological problem, which was also detected for the
Portuguese and the Spanish data, is the lack of a coherent structure of the professional
trajectory data where minor changes in the contract or minor activities (individual pay
rise) were mixed with major changes like changing position. Like in the Spanish case,
the exploratory analysis focused on international visits, most of which had duration of
up to 3 months. The preference for the USA as destination for Norwegian researchers
stands out very clearly from the data.
Bibliometric information is automatically updated in FRIDA, which is connected to the
major databases. The implementation of standard identifiers for organizations and
people in a country where register-based statistical information is the norm would allow
for an automatic updating of other types of information, such as job changes, PhD
degrees, etc.
Spain
Spain has a long history of standardisation of curricular information of researchers
working for the public sector. The country is also currently building a national e-CV
information system, as a result of integrating the various databases and systems that
were implemented in different universities and regions since 2001. The duplication and
dispersion of information and the high burden that researchers had to cope with when
filling in very detailed information in different platforms encouraged the Spanish
National Foundation for Science and Technology to launch a project to develop an
XML communication standard called CVN (normalised curriculum vitae). In May
2009, 25% of the researcher population was covered by this standard through the
different interconnected systems and databases. As opposed to the Portuguese case,
none of this curricular information is made available to the general public. Out of the
existing databases, the EURO-CV team picked the Scientific Information System of
Andalucía (SICA) to conduct the methodological and exploratory study for the project
as it was the pioneer system in the country, is highly reliable and covers 100% of the
public sector researcher population in the region. Data on professional career trajectory
and international visits (along with demographic data) was provided by SICA (within
the framework of a collaboration and confidentiality agreement). The information
recorded about professional trajectory and the data collection categories show problems
very similar to the Norwegian case (lack of coherent structure). The manual cleaning of
this data would have been very time consuming. We focused therefore on the analysis
of international visits. Data on duration and destinations of visits was clearly structured
and easily analysable. We studied the population that had declared visits over their
career (22% of total regional population, 6,481 researchers; 19,743 visits). The results
7
obtained show interesting patterns such as a sustained increase of visits per researcher
over time, a decrease in gender imbalances, and an emergence of new destinations (e.g.
Argentina, Mexico, Morocco) competing with more the traditional destinations (e.g.
UK, USA). Additionally clear differences in visiting patterns (in terms of duration,
frequency of visits, destination, and evolution over time) are found across research
disciplines.
Portugal
There are currently two main e-CV databases in Portugal: the FCT Information System
and the DeGois Platform, both of which have been promoted by the Portuguese
Ministry of Science, Technology and Higher Education. Since the DeGois Platform will
prevail in the medium-long term, the EURO-CV team focused mainly on this database.
The Platform is the result of the adaptation to the Portuguese national requirements of
the Brazilian LATTES platform. The portal allows for the individual management of
curricular information, the visualisation of national S&T indicators and the search for
curricula according to content related queries. It offers a vast amount of information
concerning career, education and scientific outputs. It is an almost fully public database,
which has led to its rejection by some researchers on the grounds of lack of
confidentiality of the CVs. The introduction of new CVs currently depends on the
individual initiative of researchers. The analysis of the data obtained from a sample
allowed the research team to identify several methodological problems (e.g.
incompleteness, lack of representativeness), most of which are due to the early
development stage of the Platform, which will be solved once the DeGois CV input
becomes mandatory for researchers applying to FCT funding. According to the analysis
performed, short-term mobility is not very likely to be reported in the DeGois CV,
unlike in Norway and Spain. Like in these former countries the extraction of the
information concerning full professional trajectories presents methodological problems
and requires a major effort in terms of manual coding work. The EURO-CV team had to
manually prepare the data to conduct a first exploratory analysis. At the time of this
report, some analysis of the sample (up to 825 PhD holders) had been performed,
including for instance the distribution by disciplines and destinations of researchers with
a PhD abroad.
----------
The work developed in the former three countries shows that e-CV national systems
may become tools with a very strong potential to follow S&T human resource activities
and careers and therefore to support policy making. The monitoring of mobility has
been very difficult up to date. These sources provide undoubtedly new ways to improve
the observation of phenomenon in the several countries with a very high coverage
(potentially 100%) of their target populations. Furthermore, at this stage they are the
only mechanism that permits us to track temporary visits, which are an important form
of mobility that contributes to shape research international spaces and networks. These
electronic systems were not initially designed for research purposes but rather for the
evaluation and diffusion of research outcomes and the management of human resources
and applicants within research organisations and funding agencies. The methodological
problems encountered when trying to analyse the ‘job mobility variables’ derive mainly
from this fact. A close collaboration between research teams and e-CV systems
8
managers and designers should however lead to big improvements in the near future as
simple re-arrangements of the information structure could lead to major advances in the
ease of use of the data. Following the EURO-CV results the Portuguese and Spanish
teams have produced methodological reports for DeGois and SICA respectively.
The Portuguese case, as well as the country’s experience within the ScienTI and
EuroCRIS networks reveals how international methodological and technical co-
operation is possible in this area and may lead to the spreading of the e-CV technologies
across countries. Many Latin-American countries are currently implementing systems
that derive originally from the LATTES Platform (e.g. Colombia, Argentina). These
experiences should be encouraging for an initiative to build an e-CV system at the
European level.
Ad-hoc CV databases: United Kingdom
Contrary to the situation in the above countries, the UK is characterised by a low level
of standardization and digitalisation of researchers’ curricular information. In addition,
even if data protection laws are not necessarily stricter than in other countries1, the
access to CV information seems to be difficult. The decentralisation and diversification
of educational and research services make it challenging to track the diverse ways of
storing and managing CVs in the different organisations. No national electronic CV
database exists. Most research organisations collect CVs through recruiting or funding
mechanisms. However, researchers usually submit their CVs in electronic pdf or printed
format. Therefore, the utilisation of CV information requires it to be manually
processed, as was also the case in the past in the three countries that now are developing
e-CV systems. The lack of standardisation and missing information were the major
barriers that the research team faced when performing the manual coding task. For
example, short visits were not coded and considered in the study due to the lack of
consistency in the information provided. Some scientists do not include the information
on visits in their CVs. The EURO-CV team in the UK had collected in the past a sample
of 173 CVs received by e-mail from scientists who replied to a CV request sent to a
population of 807 scientists and engineers. Like studies conducted in other countries on
the basis of manually coded CVs (e.g. USA, France, Spain, Switzerland) the
information was extracted from the CVs in order to study job mobility, career
progression and the influence of mobility on scientific outcomes (which turned out to be
positive). Like in the Spanish case, this study finds significant differences in mobility
patterns across the analysed scientific disciplines.
France, Israel, Switzerland
The situation in these three countries is similar to that in the UK except in that a
previously existing ad-hoc database was not available to conduct EURO-CV oriented
exploratory studies. The research systems in these countries are characterised by a
generalised lack of centralised standardization and storage of electronic researchers’
CVs. However, despite this, many higher education and research institutions collect,
store and upload CVs to the web in different ways. The Israeli and Swiss teams
1
For example, the access to CV data by our research team in Spain was supervised by the
National Agency for Data Protection.
9
provided a systematic mapping of CV collection initiatives and practices in a wide
range of both countries’ research organisations. The Swiss contribution provides a list
of all universities that have a central CV database in the country. Despite the lack of
current coordination between organisations the persons contacted in Switzerland in the
framework of the project were receptive and interested on the idea of developing a
university-wide database including a CV module. In Israel the situation is very similar.
An initiative to develop a centralised database failed in the past due to the lack of
coordination between universities and to the lack of funding. The Israel and Switzerland
sections of this report include a methodological assessment of the coding of small
samples of CVs obtained mainly through university and research centres websites.
The French contribution points out the lack of centralisation and harmonisation in CV
information collection in the country and refers to the work by Sabatier, Carrere and
Mangematin (2006)2 as a past example of ad-hoc CV database development in order to
study researchers’ careers.
European Datasources
An initial assessment of EURES and CEDEFOP European CV standards was
performed. CEDEFOP is not a database; it simply provides a general CV format, which
does not target researchers’ specifically. EURES CV Online is a job-placement service
that can be considered as an enlargement of the European Mobility Portal since it targets
all people looking for a job, not only researchers. The way in which the CV information
is stored and managed was not found out in the framework of the project. The budget
and time frame of EURO-CV did not allow for a systematic assessment of European
CV registers. In order to progress towards the development of international CV
indicators it would however be recommendable to explore in depth the potential of the
curricular information collected through the European Mobility Portal and the
scholarships and projects funding mechanisms, such as the Marie Curie program for
example.
----------
As pointed out above, the results obtained in the three countries where centralised
electronic CV systems exist show the potential that these data collection mechanisms
may have in the near future. A close cooperation between the users of the data
(researchers, social scientists, policy makers, evaluators) and the systems’ managers and
designers is essential to assure that the information is collected in a sensible and
efficient way and that the current methodological problems are overcome. Although
their origin may be linked to local initiatives, the successful development of these
systems is mainly the result of the political will of national funding or evaluation
agencies, independently of the level of decentralisation of the research system (e.g.
Spain has a highly decentralised research system and an increasingly efficient e-CV
collection and management technology). The data they store allow the analysis of large
populations of researchers, scientific production, research visits etc. and should be
2
Sabatier, M. Carrere, M. and Mangematin, V. (2006) Profiles of academic activities and
careers: does gender matter? An analysis based on French life scientists’ CVs. Journal of
Technology Transfer vol.31: 311-324
10
indeed considered when assessing S&T human resources indicators. The spreading of
the Brazilian LATTES concept and technology to a large number of Latin-American
countries and to Portugal should be looked at as a benchmark in Europe. Technology is
currently available to construct inter-connected electronic systems capable of self-
updating to some extent (e.g. information on publications). The Spanish example shows
how the development of such a system is not incompatible with preserving the
confidentiality of CV data. The obstacles to the spreading of these technologies are
nowadays mainly cultural and come from the resistance to standardization and to
privacy concerns. In order to overcome these barriers it is necessary that the policy
making and the researcher communities perceive the benefits of investing human and
economic resources in developing these types of systems. To further this, a significant
analytical effort is needed to exploit in depth the potential of the already existing
national databases and to start constructing international comparisons on the basis of the
currently available information.
2. Development of electronic CV information systems. Analysis of their
potential for researchers careers and mobility studies
2.1. NORWAY
Anders Ekeland, NIFUSTEP
2.1.1. Assessment of databases (work package 1)
The situation in Norway is marked by being in an early phase of building up a national
system of information about public research called the Norwegian Science Index,
(Norsk vitenskapsindeks). This was the primary recommendation from a working group
that the ministry of education and research established with representatives from various
stakeholders in the higher education. The working group started its work in February
2008. The report from the working group was published in October 2008 and contained
an outline of a project called “Norwegian Science Index” which in the coming years
will be constructed to take care of the need to have a knowledge base about Norwegian
public research activities. The plan for the Science Index included a CV module and a
first version was launched in the Spring 2009.
From local initiatives to multi-institution systems
The situation in Norway regarding eCVs, publication data and other kind of information
on research activities over the last fifteen years has been marked by a gradually reduced
fragmentation. The rapid technological development in the early eighties – and the lack
of inter-institution networking - led to several local initiatives at the universities and
state colleges.
This process seems to be originally driven by librarians who saw the new possibilities
offered by ICTs in their work with publications and other research related information.
The publication oriented starting point has meant that eCV applications did not emerged
from these librarian driven efforts. Typically the only CV programme that has been
operative so far was made by a student with support from the management at one of the
11
regional, i.e. university level high schools, where the rationalisation of the staffs CVs is
much more of a daily concern. This program “Kompetansetjenesten”3 or rather the
experiences were then taken over by the BIBSYS, the national agency for academic and
scientific libraries. BIBSYS then decided not to use the software but write a new one
from scratch. A beta version was launched in May 2008, but it will not be developed
further into a complete system for reasons discussed below.
Output based funding – a major driver of development of the Science Index
The gradual introduction of “result based core funding” from 2000 onwards for the
Universities and State Colleges has of course been a major driving force for getting one
integrated system in order to have the same type of data with the same type of quality
control in all parts of the higher education system that also are involved in research. In
addition “result based core funding” has been introduced for the “institute sector”, i.e.
the former public, now private non-profit research institutes from 2009. But although
there does not seem to be any immediate plan to integrate the research institutes into the
Norwegian Science Index, that will probably happen at some point in the future.
The first attempt
The development of this kind of research information management system started over
20 years ago. Starting with local initiatives – then through a gradual process – most
systems fading away, some systems merging - there emerged two systems that contain
most of the information you would like to have in an eCV system. The first is the one
called BIBSYS Forskdok (BIBSYS Research Documentation). This is the system based
on the regional high schools. The other is Frida (Forsknings Resultater, Informasjon og
Dokumentasjon av vitenskapelige Aktiviteter, Research results, information and
documentation of scientific activities) supported by the established universities4.
Neither of these systems has – or had - CV information (career) information as their
main raison d’être. Both started out from overviews of research publications. There was
and are efforts to integrate project data, but I will not describe the attempts to register
such type of data in this report. The coverage both over time and cross-section is still
rather fragmentary5. But since the share of various funding sources (research council,
ministries, business and EU) will be used as indicators, which in their turn influence the
amount of core funding there will be much stronger incentives to improve the quality
and harmonize these data.
But it was clearly wastefull to have two systems and since it is the Ministry of
Education and Research that is funding both systems – to merge them into a common
Norwegian Science Index (NSI) was the natural endpoint of the merger process, which
then started in 2009.
3
http://www.kompetansetjenesten.net/
4
There are both a paper and a presentation of the Frida system from the 2008 euroCRIS (Current
Research Information Systems, conference, see http://www.eurocris.org/public/events/conferences/cris-
2008/, paper and presentation by Grete Christina Lingjærde & Andora Sjøgren.
5
My hypothesis is that it is first when the research institutions accounting databases and/or research
project databases can deliver structured, standard, machine-readable data that can be imported routinely
as is done with the publication data that project data really will get the needed quality.
12
The structure of the two systems
The starting point for the NSI is two well-structured information systems. Both systems
were trying to meet the demand for an eCV-module. Since it has been decided to
discontinue the BIBSYS eCV-module development, this report will focus on Frida, not
discussing in any detail the experiences gained with the BIBSYS-beta version of an
eCV-module6.
Computer based creation of eCVs
We are very used to the thought that we have to write and maintain the CV ourselves,
but in since a lot of the information needed (f.ex.) publications is in a database already –
the process will probably be more and more “automatic”. When a journal accepts a
paper, the journals database will send a message to your eCV database asking it to
update your eCV with a “forthcoming” item. When the article is published, the journals
database system will ask your eCV system to update that item to a “published” item.
These kind of machine-to-machine transactions of course demands a high level of
standardisation of data and elaborate “protocols” for the transactions. But once in place
they save a lot of labour and increase the overall quality of information.
One concrete example on how registers and machine-to-machine transactions replace a
“manual” process is the census. In most countries the census is a huge survey, but for
example in Denmark they only use registers since all the relevant information like
where people live, what they earn, where they work, what education they have, family
relations etc. is already in various public registers. Consequently to construct census
data from these registers – for example when reporting to international organisations
like the EU or United Nations – is very cost efficient and you have updated information
every year. There is no reason to do a census each decade.
Therefore it essential to fully realise the enormous potential and importance of using
standard identifiers for various entities like persons, institutions, books, articles etc.
When such identifiers are used all the “digital traces” that social processes – including
scientific activity - leave behind can in principle be linked. That means that when one
has a political and/or research question one can get a lot of high quality, detailed data
about the social process of interest.
This process of using unique, standardized, numerical identifies is rapidly spreading.
The use of ISBN, ISSN, or DOIs (digital object identifiers) to identify books, articles
etc. is an example of this. A register based statistical system is a system that consciously
uses such identifiers.
When you have a policy or research question and access to a register based statistical
system you ask yourself – what kind of digital traces did the social process in question
leave behind. When the location and owner of these “traces” (data) is found - you will
then try to link together as many data sources as possible – private or public. You do not
look for “A” dataset” in singular with capital A, since in principle all data are part of the
register data system, that forms a society-wide “virtual distributed database” consisting
of all data that society records in electronic form.
6
BIBSYS took over the eCV-module ”Kompetansetjenesten”, but decided not to use the actual code in
their eCV-module.
13
The bottom line is that in principle you do not need an eCV module – the eCV can be
constructed on the basis of various registers. This is in fact already partially happening –
since both in BIBSYS and FRIDA all known publications of a researcher are
automatically fed into the eCV module. The same happens with changes in position,
from post-doc to tenure, from assistant to full professor etc.
The person is the natural statistical unit in a scientific information system. The
Norwegian standard person identifier number is used in Frida. Foreigners that have a
permission to stay also have a numerical identifier.
All institutions have unique identifiers. In Frida institutions are registered in a four level
organisational hierarchy. On top there is the “University”, then “Faculty”, then
“Institute” and at the lowest level the “research group”. This is useful for making
statistics. It also is of help to follow groups of researchers when Universities reorganise
their units, that is one can keep track of the various research groups – even if they
change institute or faculty affiliation. But so far the question of keeping track of
scientific activity- which often happens in small groups/teams – when there is
reorganisations issues are so far not explicitly dealt with. But at least the four level
hierarchy of organisational units makes this possible.
Identifiers for foreign institutions
One very important part of institutional register in Frida is the list of foreign research
institutions. The list contains more than 20.000 entities all over the world and
continuously growing. The foreign institutions are only at the highest “University”
level. It would be very good if one could know to what faculty/institute/group the
Norwegian researchers go when they have longer or shorter stays abroad.
Such a list of institutions should be a common, standardised list with agreed procedures
for updating etc. There should also be developed a common nomenclature for
describing the institutions – along several dimensions. This would greatly facilitate
research on international scientific cooperation like co-publication, mobility etc.
As mentioned above – the automatic feeding in of bibliometric data from external
bibliometric databases Norwegian, Nordic and ISI is an important example of the
usefulness of “register data”. The alternative is that the researchers themselves write
publication data into the system. Not only is this a waste of scientific labour, but it also
means non-standard writing of names etc. etc. – reducing the quality of the data.
Since publication data are still not fully standardised there is a sophisticated user-
interface related to registering and/or importing publication data in order to ensure
uniqueness and consistency. More sources are being gradually included, wrong/disputed
classifications corrected etc. For the study of research productivity these are very
valuable data.
To sum up, the core issue is to use standard identifiers, to have conscious data structures
(conceptual hierachies) – then eCVs can be constructed from data in various domains
like publication, visits, positions etc. For many years to come eCVs will be “manual” or
at best semi-automatic; but one should always strive for further standardisation and
machine-to-machine integration.
14
Classification – aggregation - conceptual hierarchies
A general problem in statistics, and also in this case is that the the nomenclatures, i.e the
code tables are generally too ”flat”. That is they have too little analytical structure. The
“position codes” (professor, ass. prof. etc.) are not easy to separate from administrative
position codes, since they do not use the first, second, third digit … in systematic
manner. This means that there is time-consuming data structuring job – writing long
lists of numbers - to do in order to be able to separate administrative positions, from
research positions. The codes for the fields of study also lack structure making it
difficult to aggregate. There should be multiple classification trees to map inter-
disciplinary work, like “climatic research”, “nano-technology” or to separate language
studies from more general area and culture studies. If data are going to be useful in an
international context such higher level classifications are needed, preferably on the basis
of an international standard.
CVs - or publications and profiling?
The career oriented CV has not been a major focus of neither Frida nor BIBSYS. As
mentioned above BIBSYS launched a first version before further development was
stopped. Frida launched a first version of its eCV module in the spring 2009.
Data from the eCV modules
To get a more detailed picture of the status of the eCV modules, I applied for and
obtained data both from BIBSYS and Frida. The BIBSYS eCV was at that time only in
beta version and contained very few data, only five hundred cases and many of them
with a lot of missing variables. Due the quality of data – and the fact that BIBSYS is not
developed further, I will only discuss the data from the Frida system.
The Frida system in December 2008 covered the following institutions.:
• University of Oslo
• University of Tromsø
• Norwegian University of Science and Technology -
• University of Bergen
• Oslo University College
• The Norwegian Institute of Public Health
• Norwegian Knowledge Centre for the Health Service
The current rate (June 2009) of coverage of university level institutions is
approximately 75%, but since Frida is chosen as the national system, coverage will be
100% in the near future.
To make the process of getting access to data from Frida easier I limited my request to
the four “old” universities (Oslo, Bergen, Trondheim and Tromsø). Tromsø had some
question relating to privacy that could not be resolved in the time left for this project so
I did not get data from Tromsø. The list of requested variables as co-ordinated with the
Spanish team, focusing on demographic information, career trajectory and temporary
research visits.
I also had to abstain from getting the person identifier to link the data I did get to other
register data. The lack of person identifier means that previous work career and longer
15
stays abroad cannot be linked in from employer-employee data and migration files. The
There are work career and migration data back to 1986.
A small aside – the Norwegian register based statistical system
This possibility to link data follows from the very wide-spread use of the standard
person identifier in Norway and the other Nordic countries. This us pre-dates the
“computer-age”, i.e. was established in the sixties in order to get unique identification in
paper-based registers. But once in place it spread to the banking system, to “civil
society” organizations as they became computerized. There was also various identifiers
used for firms and organizations, one by Statistics Norway, another by the Social
Security Agency, the tax authorities, but in 1995 there was established a common
register for all firms and organizations (state, municipal organizations/institutions,
political parties, sports clubs etc.). Since for example the education institutions use the
person identifiers – all exams a person has taken can be collected by Statistics Norway
on a machine-to-machine basis. In this way you can (slowly) build up an integrated,
register based statistical system – although it was not planned from the start. The key
issue is that every organization uses standard identifiers – person identifier,
firm/organization identifier – since then these data can be linked together at a later
stage. It is a myth that such a system has to be build as an in advance centrally
controlled system.
The resulting dataset
The resulting dataset had a bit more than 16.000 persons with unique identifiers, but if
we disregard guests and prof.emerit we are down to a core population of 12000. The
bulk of the population is of course the universities of Oslo (5000), Bergen (3800) and
Trondheim (3200). These numbers must just be seen as a rough indicator, since there
obviously are some differences in the way employees are registered in the universities
own personnel databases – and the routines for export to Frida.
No system is perfect from the start and there are some minor problems with the
population in Frida. The first is that guests and associated persons such as visiting
researchers, professors emeritus, etc, are not always registered in the local personnel
system. The ad hoc solution has been to manually register these persons, but one could
think of using other registers (entrance cards registers, list of registered users of the
University network) to check for such persons. The second problem is the interface
between the administrative systems and Frida. In the data that I got there was clearly a
non-optimal structure, all minor changes in the contract (individual pay rise) were
mixed with major events like changing position, changing faculty etc. – since this is the
way that the data are registered in the personnel database of some of the universities.
The various types of career changes need to be more carefully classified.
The above mentioned minor problems with the data, i.e. no previous career, major and
minor changes in the contract lumped together could have been overcome. Previous
work history could have been integrated by linking with other register data giving the
full career of workplaces, but not positions outside the academic institutions. The major
and minor contract changes could have been separated by some tedious data massage.
The degree of massage one would have to do would depend on the precise analytical
questions.
16
2.1.2. Work package 2 (exploratory studies): Data on scientific stays abroad
The data finally obtained – regrettably only from the University of Oslo – was data on
short term scientific stays abroad. I will not attempt any real analysis here, just a short
description of the data and their potential for future attempts to map and measure
international migration (long term stays) and scientific networks. There are totally
around 8200 persons in the database not classified as guests since, some 200 - 500 are
probably more administrative oriented positions. In comparison there are registered
4000 “guests” totally from 2002 onwards. So it is possible to study also the
immigration, but as far as I know there is no data on the institutions where visitors are
coming from. This would of course be very valuable data to see the amount of exchange
visits.
The figure below shows the number of visits abroad per person. It could of course be
broken down by the available background variables, like gender, ages, discipline,
position etc. The unit of analysis is the individual researcher. In addition to data on the
visits there are of course a lot of other data available about the individual, gender, age,
education, previous academic positions, publications, etc.
Figure 1: Frequency of visits abroad per person, scientific staff only.
500
450
400
350
300
250 Freq
200
150
100
50
0
1 2 3 4 5 6 7 8 9
N= 1144, Source: Frida
As expected most people have only one stay abroad per year, but two is not that rare. If
we look at the destinations it is no surprise that US comes out on top.
17
Figure 2: Frequency of destinations – all visits
300
250
200
150
100
50
0
ia
na
r-A e
Et s
A
F r and
ia
Sv i a
d
e
Fi a
us a
ia
do ia
e
an a
ed ark
Ty ia
ka
Au n
lia
d
d
a
n
t
Sø rig
ny
ik
pi
Zi lan
bw
ad
an
an
si
le
pa
or US
rk
ei
an
l
lg
an
nn
Ita
ra
fri
Ki
kr
ne
io
Po
m
Sv
Be
Ty
Ke
e
nl
l
Ja
an
sl
ba
sk
er
Sp
st
nz
ita
an
C
br
m
Ta
D
R
In
N
St
N= 1144, Source: Frida
Figure 3: Frequency of scientific visits abroad in 2005, 2006 and 2007
100
90
80
70
60
50 2005
40 2006
30 2007
20
10
0
ge
ts
SA
na
d
e
ia
e
ka
a
a
lia
n
lia
rik
n
ni
ig
pi
pa
nn
ei
Ita
ra
fri
or
Ki
la
U
er
a
io
Sv
nk
Ja
r-A
ita
sk
Sp
st
N
Et
Sv
a
Au
Ty
br
Sø
Fr
or
St
N= 1144, Source: Frida
Although being based on only three years it seems that the numbers of visits are fairly
stable. The exception is China, which seems to get up to a new level.
18
The distribution of length of stays has maybe shorter – one or two weeks stays – and
less medium – 2-3 months stays than expected.
Figure 4: Length of stay abroad
350
300
250
200
150
100
50
0
A week Two weeks A month 1-3 months 4-6 months 7-12 months More than 12
months
2.1.3. Conclusions and recommendations
Regarding eCVs in Norway the eCV part of Frida has been launched, it is still not
mandatory to use when applying for funding, it is still not a social norm to use it, so to
what extent this module will be used by Norwegian researchers as their CV-tool
remains to bee seen.
But the Frida – and in the future Norwegian Science Index will contain most of the
information that one expects to find in an eCV, so that for research purposes – and with
the data from the other part of the Norwegian register data system – the data needed for
research purposes should be available.
The main points from the Norwegian experience so far are:
• The importance to use standard identifiers for organizations and people, i.e.
person identifiers, establishment and enterprise numbers. This is only partly in
place
• To think analytically from the start, that is policy makers and researchers should
discuss the classifications needed in order to meet present and future needs.
• One should in particular look at “the demand side” – i.e. to demand a certain
standardization of CVs when applying for funding etc. There is no reason not to
have a CV standard also used on the general labour market7.
7
This is fact emerging, since all the ministries and f.ex. Statisics Norway use the same eCV database for
people applying for jobs.
19
• The development of eCV should be closely connected to the efforts of
registering other kind of research input and output data. Projects, conference
participation etc – not only books and publications in peer review journals.
• There is an urgent need for a common numerical, classified register of education
and research institutions on a European/international basis in order to study the
short and long term researcher mobility. We need to know not only the country
they go to, but the institutions visited
• There is an urgent need to get and international numerical scientific author
register – and connected to the DOIs (digital object identifier) of publications.
This would do away with a lot of the “manual” name-matching in bibliometric
and patent research.
References
Lingjærde, G C and A Sjøgren 2008. Quality Assurance in the Research Documentation
System FRIDA. Paper presented to the 9th international Conference on Current
Research Information Systems, 5–7 June, Maribor, Slovenia. <http://www.
eurocris.org/public/events/conferences/cris-2008/>
2.2. SPAIN
Carolina Cañibano, Javier Otamendi, Inés Andújar; Universidad Rey Juan Carlos; Félix
Muñoz, Universidad Autónoma de Madrid.
2.2.1. Assessment of databases (work package 1): Introduction to the eCV system
in Spain
The Spanish system of science and research is characterized by two important features
as far as the use of researchers’ CVs is concerned: standardisation and increasing
digitalisation. There is a tradition of CVs standardisation in the country as public
funding agencies require researchers to present their curricular information in
standardised formats, which may only vary slightly according to programmes. CV
standards permit organisations to improve the administrative processes, such as the
evaluation and management of grants and funding support.
Increased digitalisation results from the expanded use of new technological tools among
public research institutions and funding agencies to collect and store researchers’ CVs.
A growing number of universities and research public organizations are using electronic
platforms which permit researchers to introduce their curricular data and organisations
to use this data for different management and evaluation purposes. A wide researcher
CV information system is under development in Spain. As a result, a considerable
amount of curricular information is stored in an electronic and standardized way.
However, this information has not been yet exploited for research purposes.
As will be elaborated later, the Spanish CV information system is formed by more than
40 organizations, including the former Spanish Ministry of Science and Education
20
(MEC)8, the National Agency of Quality Evaluation and Accreditation (ANECA),
universities and other public organisations (such as the Scientific Computing Centre of
Andalucía, CICA). Each organisation stores the CVs through electronic systems of data
management. The number of the collected CVs varies from 40.000 in the ANECA or
20.000 in CICA to 700 or 400 in certain universities.
It is important to emphasize the computer systems’ diversity (around 14 different
systems are being used). For example, CICA uses the SICA system; ANECA - the
ANECA-CV; the MEC - the SIC-MEC. This creates barriers for the curricular
information circulation, with the corresponding lost of efficiency. Moreover, the
different computing systems have diverse scopes of application: national (in case of the
MEC and ANECA), regional (SICA) or within the university scope. A researcher has
therefore to provide the same information in different computer applications for
different organizations.
A great progress has recently been made with the ongoing implementation of the CVN-
XML system (Normalized Curriculum Vitae Project)9 undertaken by FECYT (Science
and Technology Spanish Foundation). The CVN is a communication standard which
allows for the integration of the existing data sources through the use of a common CV-
XML format. Further details concerning the CVN project are provided later in this
report.
The Curricular Information System (CIS) is a result of databases interaction, which does
not however imply the creation of a common database but the intercommunication of
the existing ones and the possibility for the automatic transmission of data. If, for
instance, a researcher updates his or her CV in a database and provides the
corresponding authorisation, the data may automatically be updated into other databases
of the system. This will presumably reduce the big bureaucratic load that researchers
need to cope with in Spain.
The new CV information system opens up new possibilities for the analysis of the
trends and patterns of the Science and Technology System in Spain, as it contains a
massive amount of data that can be easily exported and analysed. The CVN-XML
standard contains up to 1200 variables, which include sociological data (age, gender,
nationality, address, etc), education (masters, PhD) job trajectory, international research
visits, grants, projects, publications, contributions to conferences, patents, supervised
thesis, teaching, etc.
According to recent data from FECYT (May 2009), the number of eCVs available in
CVN format is 30,440, which represent 25% of the total researcher population in
Spain10. Overall, the system contained 106,340 CVs in May 2008 (in different formats,
as the CVN format was not in use yet); this number over-rates the coverage of the
system, as it does not account for duplicated CVs. The introduction of the CVN
communication standard is permitting the system to avoid and control duplications.
8
The MEC has recently been divided into 2 Ministries: The Ministry for Education, Social Policy and
Sport and the Ministry for Science and Innovation (MICINN). The new MICINN is responsible for the
management of the CV platform.
9
For additional information of the CVN Project visit:
http://cv.normalizado.org/presentacion/institucional.jps
10
Rate calculated on the basis of the Spanish R&D survey data 2007 (www.ine.es).
21
Table 1 bellow provides the list of institutions which are using CV information systems.
Table 2 presents the list of different databases and systems. Out of the existing
databases and systems, 1 has finally been selected for exploitation and methodological
assessment within the EURO-CV project: SICA. The system is described in detail in
section 2.2.3. A brief description of the rest of systems is provided bellow.
Table 1: Institutions using CV electronic information systems in Spain
CV managment Number of
Acronym Institution Name Web Scope
system researchers
ANDALUCIA Junta de Andalucía www.andalucia.es SICA 20000 regional
Agencia Nacional de
Evaluación de la Calidad y
ANECA Acreditación www.aneca.es ANECA-CV 40000 national
www.goviernodecanar
CANARIAS Gobierno de Canarias ias.org CANARIAS CV n/d regional
Centro de Supercomputación
CESGA de Galicia www.cesga.es RIGA n/d regional
DEUSTO Universidad de Deusto www.deusto.es OCU 800 University
EHU Universidad País Vasco www.ehu.es IKERTU 4500 University
Extremadura Universidad de Extremadura www.unex.es OCU 1200 University
Instituto Catalán de
ICO Oncología www.scs.es GREC n/a Research institute
JCYL Junta de Castilla y León www.jcyl.es NOVATORES n/a regional
Ministerio de Educación y
MEC Ciencias www.mec.es SIC-MEC n/a national
CAMPUS
UA Universidad de Alicante www.ua.es VIRTUAL n/a University
Universitat Autònoma de
UAB Barcelona www.uab.cat FENIX 3400 University
UAH Universidad de Alcalá www.uah.es OCU 1300 University
UB Universitat de Barcelona www.ub.edu GREC 6150 University
UBU Universidad de Burgos www.ubu.es OCU 700 University
UC3M Universidad Carlos III www.uc3m.es OCU 1500 University
Universidad de Castilla la UCLM
UCLM Mancha www.uclm.es Investigación 2300 University
UDC Universidad de la Coruña www.udc.es RIGA 1300 University
UdG Universitat de Girona www.udg.es GREC 2442 University
UdL Universitat de Lleida www.udl.es GREC 1216 University
Universitat de les Illes
UIB Balears www.uib.es GREC 1259 University
UJAEN Universidad de Jaén www.ujaen.es OCU 900 University
UM Universidad de Murcia www.um.es GINVEST n/d University
UNAV Universidad de Navarra www.unav.es OCU 1300 University
Universidad Pública de
UNAVARRA Navarra www.upn.es FENIX n/d University
UNICAN Universidad de cantabria www.unican.es CV UNICAN 0 University
UNILEON Universidad de León www.unileon.es OCU 800 University
UNIRIOJA Universidad de la Rioja www.unirioja.es OCU 700 University
Universitat Oberta de
UOC Catalunya www.uoc.edu OCU 700 University
Universitat Politècnica de
UPC Catalunya www.upc.es FENIX n/a University
UPF Universitat Pompeu Fabra www.upf.es FENIX n/a University
Universidad Politécnica de
UPV Valencia www.upv.es CARTA n/a University
URJC Universidad Rey Juan Carlos www.urjc.es OCU 800 University
URL Universitat Ramon Llul www.urv.es GREC 425 University
URV Universitat Rovira i Virgili www.urv.es GREC 1706 University
US Universidad de Sevilla www.us.es SICA n/d University
USAL Universidad de Salamanca www.usal.es OCU 2300 University
Universidad de Santiago de
USC Compostiela www.usc.es RIGA 2470 University
UV Universitat de Valencia www.uv.es GREC 3172 University
UVA Universidad de Valladolid www.uva.es OCU 3000 University
UVIGO Universidad de Vigo www.uvigo.es RIGA n/a University
Source: FECYT, May 2009
22
Table 2: CV electronic information systems in Spain
System Name Información Adicional
System Acronym
Sistema gestor de CV de la ANECA
ANECA-CV
Sistema Gestor de la Universidad de Alicante
CAMPUS VIRTUAL
Oficina de Ciencia y Tecnología del Gobierno System under development. Does
Canar not contain any CVs yet
CANARIAS-CV
Sistema Gestor de CV de la UPV
CARTA
Sistema Gestor de la Universidad de System under development. Does
Cantabria not contain any CVs yet
CV UNICAN
FENIX Sist. Gestor desarrollado por la
FENIX empresa SIGMA
Universidad de Murcia dentro del módulo
GINVEST Ática
Aplicación de Gestió de la Recerca. Liderado
GREC por la UB
Sistema Gestor Universidad País Vasco
IKERTU
Sistema gestor de CV de Castilla y León.
NOVATORES Lidera la USAL
Empresa que hace sistemas gestores de CV,
OCEANOGRAFICA Canarias
Universitas XXI investigación
OCU
Rexistro de Investigadores de Galicia
RIGA
Sistema de Información Científica de
SICA Andalucía
Sistema de Información Curricular del System under development. Does
SIC-MEC Ministerio de Educación y Ciencia not contain any CVs yet
Sistema Gestor de la UCLM
UCLM Investigación
Source: FECYT, May 2008
• ANECA-CV: The system manages approximately 40.000CVs (information from
May, 2008). It is used and managed by the National Agency of Quality
Evaluation and Accreditation in Spain (Agencia Nacional de Evaluación de la
Calidad y Acreditación) to perform its statutory functions. The agency was
created in 2002 to contribute to the quality improvement of the higher education
system through the assessment, and accreditation of university degrees,
programmes, teaching and research staff and institutions. Researchers submit
their electronic CVs through a CV platform in order to apply for evaluation.
• CAMPUS VIRTUAL: The system is used and managed by the University of
Alicante. It is a platform dedicated to students, teaching and research staff and
administrators of the University. The platform was created to facilitate the
administrative work and communication inside the institution.
Using the CV application, researchers can manage their CVs, see their wages,
interact with members of their research group and consult the outputs and
finances of their research projects. The system uses Internet and the HTML as an
interface, but the CV can also be generated in the RTF format.
23
• CANARIAS-CV: The system is used by the Government of Canarias and is
managed by the regional Office of Science, Technology and Innovation. It was
developed to facilitate the performance of the statutory functions of the Office,
i.e. enable communication between the companies, public and private research
centres, to plan and coordinate the participation of Canarian researchers in
international research programmes etc.
• CARTA: The CARTA system (Catálogo Corporativo de Capacidades y
Resultados Tecnológicos y Artísticos) is used and managed by the Universidad
Politecnica de Valencia (Polytechnic University of Valencia - UPV). It can be
accessed by companies and institutions from the private and public sector that
wish to exploit UPV’s research outputs (patents, licenses and other technology)
applicable to business. The system uses KNS (Knext) which is a Web
application and complies with accepted Internet standards (XHTML, CSS,
XML).
• CV UNICAN: The CV UNICAN is currently under development and does not
contain CVs yet. It will be used by the University of Cantabria.
• FENIX: The FENIX system manages more than 3400 CVs in 4 universities. It is
managed by the Spanish company Sigma Gestión Universitaria AIE that
integrates seven universities and is devoted to creation and implementation of
software for the management of universities. The system allows researchers to
place their CVs on the university’s website and extract it and to consult the
outputs and finances of their research projects. The SIGMA software is operated
via Internet and it is constructed in Java and J2EE.
• GINVEST: The system is used by the University of Murcia and is managed by
the ATICA (Área de Tecnologías de la Información y las Comunicaciones
Aplicadas), a research centre of the University created in 2001. GINVEST
integrates the management of ongoing PhD thesis, fellowships, research
projects, patents and other scientific output. It supports the creation and follow-
up of research groups. The system allows researchers to add their CVs to the
data base. The University of Murcia, uses the data contained in GINVEST to
write its research Annual Report. The system uses Internet and the HTML as an
interface.
• GREC: The system contains more than 16.000 CVs from eight Spanish
universities in Cataluña. It is managed by the Gestió de la Recerca Group with
the University of Barcelona as the leader of the Group. It is an Internet platform
created to facilitate managing of the scientific activities at universities. The
GREC system allows managing research grants and projects and their outputs.
The Curricul@ application permits the organisations to use CV data to generate
research reports and output indicators for evaluation purposes.
• IKERTU: The application manages approximately 4500 CVs. It is used and
managed by the University of the Basque Country. It supports human resources
and research projects management. The user can create the necessary reports that
are used in the decision-making process at the university.
• NOVATORES: It is the researcher CV database of the Castilla y León region.
The NOVATORES platform is mainly administrated by the University of
Salamanca. It was created in 2003 in order to use Internet to diffuse information
on the scientific output produced in the region. The platform was developed to
be used by researchers, scientific institutions and companies of the Castilla y
León region. The CV application allows the researcher to maintain his CV
always updated and to apply for the fellowship or grants.
24
• RIGA: The RIGA system (Register of the Researchers of Galicia) contains more
than 3700 CVs. It is managed by the Junta de Galicia Assembly and it is an
Internet application. It allows for the collection of personal and professional data
of researchers. The CVs submitted through the platform can be used to apply for
grants funded by the Junta de Galicia.
• SIC-CINN (Sistema de Información Curricular del Ministerio de Ciencia e
Innovación; previous name: SIC-MEC): It is a system created and managed by
the Spanish Ministry of Science and Innovation, however, at the moment it is
still under construction. The platform is the Ministry’s contribution to the
implementation of the CVN norm of CV standardisation. Once the registration
process is completed, the researcher will be able to fill in the CV template
constructed according to the CVN norm, save it in the PDF or XML format,
create a summary of the CV or personalize it according to the needs of the
particular application.
• UCLM Investigación: The application contains approximately 2300 CVs. It is
used and managed by the Research Office of the University of Castilla la
Mancha.
• UNIVERSTAS XXI: This platform is used by several public universities. The
research module includes a CV platform. The system allows researchers to save
and print their CVs in different formats depending on their needs. The option to
generate the CV in CVN-XML format has been very recently incorporated to the
system. The EURO-CV team from the Universidad Rey Juan Carlos (URJC)
obtained permission to access the contents of the Universitas XXI curricular
data-base at the University. The URJC data base contains information on 3411,
out of which around 1700 had submitted their completed CV in May 2008. The
downloading of the curricular information is however very time consuming and
requires much manual coding, as the system was not designed to allow the
conduction of this type of study. For this reason, we finally decided not to use
this data source to conduct the exploratory study.
2.2.3. The scientific information system in Andalusia: SICA
SICA is the pioneer researcher curricular information system in Spain. It was created as
a support for the regional science and technology system of Andalusia in 2001. It is a
repository of curricular information from researchers either working in Andalusia or
cooperating with Andalusian researchers. This information can be used by researchers
themselves, evaluators, research organisations administration and policy makers.
The curricular information contained in the system can be transferred to other
institutions for different purposes, especially for evaluation and grant and projects
selection processes. The tool helps evaluators by providing them with the information
they specifically demand concerning the researchers or projects under evaluation.
SICA is managed by CICA (Andalusian Scientific Information Centre). A very
interesting specificity of this system as that it is automatically updated through its
connection with major publication databases such as ISSI, SCOPUS, CINDOC and
MEDLINE. Moreover, the SICA personnel are in charge of validating the information
that researchers introduce in the system by checking the contents of the corresponding
databases. SICA is therefore probably the most reliable curricular information system in
25
the country at this stage. All Andalusian researchers CVs are currently registered in the
database, which also contains CVs of researchers who collaborate (through projects,
grants, contracts) with the regional system. SICA currently contains more than 30,000
CVs.
As other systems SICA is currently implementing the use of the CVN-XML format but
it already provides the possibility to export CV data in XML, XLS or MDB formats.
The URJC EURO-CV team signed a cooperation agreement with SICA in order to
obtain curricula and conduct exploratory studies. The characteristics of the data
obtained will be described in section 2.2.5.
2.2.4. The CVN-XML standard: Towards a normalized electronic CV
As discussed earlier the standard CVN-XML allows research organisations with
different curricular information systems to be interconnected and exchange information.
The CVN project was launched in 2006 by the Spanish Foundation for Science and
Technology (FECYT).
The shared use of the CVN is intended to become a basic instrument for the exchange of
curricular information with the aim of facilitating the management of public and private
R&D grants and projects in all organisations within the Spanish Science Technology
and Enterprise System (SECTE).
The CVN-XML is a standard communication system which permits the existing CV
databases (see Table 6) to be connected with each other and share information. The
standard CV includes so far 567 items which are divided into 7 chapters. Thanks to the
CVN research institutions (universities, national and regional evaluation agencies) will
be able to share and exchange information: i.e. when a researcher applies for funding to
one of them and updates his or her CV, all of them can have access to it. Moreover, the
aim is that all of the SECTE institutions may administrate the same format of
information. The CVN format is standardised and, theoretically, multilingual.
The contents of the CV
The CVN model has all the information that all classes of researchers and research
technicians may need in order to produce an outline of their work and results, including
their interaction with other academic staff or sectors.
The CVN standard includes more descriptions than one person would generally use.
Nevertheless, it is beneficial to include different types of work in a single format since
various researchers may perform various types of activities simultaneously. Each person
uses the needed fields depending on the career stage and on the institution to which the
CV will been sent to.
In May 2009 the total number of CVs available in CVN format in the system was
30,440, which represents 25% of the total FTE Spanish researcher population. The CVN
norm is currently in use at 13 universities and in the entire Andalucía region.
26
Figure 5: The CVN information system structure
2.2.5. Exploratory studies (work package 2)
Once the national curricular information system was characterised, the research focused
on the conduction of the corresponding exploratory study in order to address the
project’s research questions. The subsystem used for this purpose was SICA, which was
the pioneer CV information system in the country and has been largely used and
improved since 2001. The information contained in the system is highly reliable and
updated and covers the entire population of researchers in the region of Andalucía.
To start with, we requested data from the sub-population of researchers working at the
University of Sevilla, which is the largest university in the region and employs more
than 7000 researchers. Our purpose was to assess how mobility and career trajectories
could be analysed on the basis of the information downloaded from the system.
We requested and obtained the following variables:
A) SOCIOLOGICAL VARIABLES
• Gender
• Country of residence
• Date of Birth or Age
• Educational level
• Professional category
27
B) RESEARCH VISITS
• Duration (beginning and closing date)
• Type of visit
• Receiving institution or organization
• Country
• City
C) PROFESSIONAL TRAJECTORY
• Duration (beginning and ending dates of the contract or activity)
• Post
• Organization
• Type of experience or activity
• Country
Among these variables, the following are particularly hard to work with for the below
indicated reasons:
• Professional category: this is an open field variable. The researcher chooses
what to indicate. Therefore, for example, different terms are use to refer to the
same type of position by different researchers. Some professional categories are
described in a very vague way, such as “scholarship holder” for example. The
use of this variable for analytical purposes would require a considerable effort of
manual cleaning. Our recommendation would be to convert the field into a
limited-choice one, so that the researcher has to choose among given options.
Additionally, an open field could be left to introduce further comments or other
eventual types of positions not included in the given list.
• Type of visit: This is a closed field variable. However, the proposed categories
are not mutually exclusive. These are: scholarship holder, invited researcher,
contracted researcher, doctoral, postdoctoral, other. The classification is not
operational for analytical purposes. Our recommendation would be to define a
set of mutually exclusive categories.
• Post (within the professional trajectory set of variables): This is again an open
field. Some researchers include only their job contracts in this section, while
other list all their professional activities, from journal refereeing, to the
organisation of conferences. Manual cleaning of this variable in order to study
job mobility would also be very time consuming. We recommend making a clear
distinction in the CV format between job contracts and other types of
professional activities.
Given the difficulties to address job mobility on the basis of the available data, we
decided to focus the exploratory study on international research visits, using all
available demographic variables plus the information on duration and destination of the
visits. The literature acknowledges the importance of temporary mobility of researchers
for the formation of collaborative networks and the diffusion of scientific knowledge
(Woolley & Turpin, 2009; De Filippo et al. 2007; Ackers, 2005). The European
Commission has started to consider the encouragement of short-term visits, called
“shuttle-stays” (CEC, 2008) as a means to support knowledge transfer to developing
28
countries. This approach seems to be widely accepted as proves the development of the
United Nations TOKTEN programme or some national recent initiatives such as the
Uruguayan11. In Spain, both regional and national governments have consistently
increased over the last two decades the amount of resources devoted to fund
international temporary research visits.
As mentioned above, our exploratory study focuses on the Andalusian population of
researchers. The Andalusian research system employs 20% of Spanish researchers. The
SICA system covers the entire population of public sector researchers in the region. In
May 2009 the system had 30,737 researchers’ CVs, most of which work at universities
(76% of total).
Out of the total number of researchers in the region, the population under study is
formed by researchers who have registered international research visits in their
electronic CVs. The total number of researchers is 6,955 and the total number of visits
20,990. Out of this total we eliminate from the study the visits longer than 2 years and
shorter than one week. We also eliminate from the study the visits with error or missing
data (for example, visits to Spain as country of destination, as this cannot be considered
international visits or visits without an ending date). The final number of studied
researchers is 6,481 with a total of 19,743 visits.
The population under study is mainly formed by researchers working at universities
(88%), with a higher rate of man (60%) compared to women researchers and a higher
proportion of researchers aged between 30 and 40 (43%). 70% of the sampled
researchers are doctors. Table 3 shows the distribution of the sample according to
discipline. The humanities count with the largest proportion of researchers in this
sample (just as in the whole population of Andalusian researchers), followed by the
social sciences and by physics, mathematics and chemistry.
Table 3: Distribution of the sample according to discipline
11
The Uruguayan Government approved in November 2008 to devote 15US$ to fund short (a few weeks)
stays in Uruguay of 1000 Uruguayan researchers residing abroad, in order to encourage knowledge
transfer.
29
International visits of Andalusian researchers:
General overview
Out of the population of Andalusian researchers, those who have registered at least one
international visit over their careers account for 22% of the total which in turn implies
that more than 3 quarters of the population have never made (or never declared in their
CVs) an international visit. Our data is not directly comparable with other studies other
studies conducted in Spain. A recent study from the European Commission targeted at
assessing the obstacles to researcher mobility in the EU found that among all surveyed
researchers (3,365), 24% were mobile (CEC, 2008). However, mobility in this study
was defined in terms of residency abroad and not of international research visits.
We do not really have a comparative basis to assess the rate of short-term mobility
found in our study. It appears however to be rather low.
Figure 6 shows the distribution of visits’ duration. Out of the total 19,743 visits, 68%
had a duration of one week to 3 months, 25,5% lasted between 3 and 12 months and
6,4% lasted between 12 and 24 months. Very short-term visits are thus the most
frequent ones.
Figure 6: Distribution of visits in terms of length
Figure 7 shows the number of visits per researcher. Many of them (37%) register only 1
visit over their careers. Most researchers (90%) declare between 1 and 6 visits.
30
Figure 7: Distribution of visits per researcher
Western Europe is the most important destination area, with 61% of registered visits,
followed by Latin-America (18%) and North-America (14%). Figure 4 shows how the
UK, France, Italy, Germany and the USA attract 58% of visits. It is interesting to point
out however that the three following main destination countries are Mexico, Portugal
and Argentina. The evolution of destination countries over time shows some interesting
features that we will comment on below.
31
Figure 8: Countries more often visited
The gender and age distribution of visits shows also some interesting patterns. If we
consider the full sample under analysis men are significantly more mobile than women
(p-value= 0.0043). However, mobility gender differences are not significant for the
younger cohorts (25 to 40 years).
Dynamic analysis
The total number of international visits has substantially increased over time, which is
mainly the result of growth in researchers’ population. We estimate the evolution in
visits per researcher dividing the total number of visits per year by the number of
researchers in the sample who were 25 or older in that same year. We consider that
researchers may start being mobile from that age, while they work on their doctoral
studies. Figure 9 shows how visits per researcher have steadily increased over time
since the mid 1980’s, which could be linked to the first policy actions to sustain
international mobility and the entry into force of the ‘Law of Science’ (Ley de la
Ciencia) in 1986. Since the mid 1980’s, the sustained public effort to support
international mobility of researchers by the national and regional governments and the
progressive internationalisation of scientific practice would certainly be explanatory
elements of the sustained trend. We associate the decrease of the most recent years
shown in the figure to the possibility that many researchers have not yet updated the
information concerning recent visits in their CVs.
32
Apart from the general trend, it is important to point out that the series reaches a
maximum of approximately one quarter of a visit per year, which implies an average of
1 visit every for years.
Figure 9: Visits per researcher and year
To study the evolution in geographical destination we classify countries in 7 major
regions. Significant changes appear: Africa and Latin-America emerge as destinations
after 2000, while the USA and Western Europe experience a relative reduction. A bi-
factorial analysis of the variables “decade” and “zone” after controlling for sub-sample
sizes confirms these tendencies. The signs in Table 4 show significant variations in the
preference for each zone compared to the others, for each decade. It is interesting to see
how since 2000 new geographical destinations have emerged while the rate of
traditional destinations has diminished.
Table 4: destinatios by decade
Figure 10 results from combining the variables “country” and “decade” in which the
visit took place. Only countries that show significant changes between decades are
shown. We must point out the emergence of Brazil, Mexico, Portugal and Morocco,
which contrast with the significant decrease for more traditional destinations such as the
UK, France or the USA.
33
Figure 10: Countries with significant changes amongst decades
Mobility patterns by discipline
Demographic characteristics of the discipline sub-samples:
34
Table 5: Summary by discipline
35
Table 5 shows how the rate of mobile men is systematically higher than that of mobile
women in all disciplinary fields. It is particularly higher in Physics, Chemistry and
Mathematics (FQM), Production Technologies (TEP) and Information and
Communication Technologies (TIC). The analysis of the overall mobile population
shows a higher tendency to move for men than for women. However, the gender
differences vary if we consider only the younger cohorts. In the age group from 20 to 30
years, the rate of women in the mobile population under study is higher in Agricultural
Sciences (AGR), Biology (BIO), Humanities (HUM) and Environmental and Natural
Resources (RNM). The same happens in the age group 30-40 for the disciplines HUM
and Social and Juridical Sciences (SEJ). These results seem thus to show a tendency
towards a higher gender balance in international temporary mobility levels.
The age distribution is quite homogenous among disciplines. The age group 30-40 is the
most numerous in all fields, followed by the group 40-50. The population in the field
RNM is significantly younger than in the rest of disciplines, which is consistent with the
youth of the discipline itself in Spain.
Finally, the distribution between doctors and no doctors shows a higher presence of
doctors in the mobile population (higher than 60% in all fields), with a significant
higher weight in SEJ (74%) and FQM (77%) and lower in TIC (63%). It is worth noting
that in the latter discipline, 37% of the mobile population is thus formed by non doctors
and that, overall, non doctors represent 30% of the population which registers visits
abroad, which shows the importance of international visits in the pre-doctoral training
period.
Mobility patterns by discipline:
HUM: Humanities (1963 researchers)
Over the studied period, research visits have increased in this discipline above the
overall growth rate of visits in the population under study. This field is characterized by
more frequent and shorter research visits than the rest of fields. The higher number of
visits per researcher and the higher frequency of visits with a duration of 1 week to 3
months stands out. Together with TIC, HUM is the discipline that registers a higher
growth of visits over the last decade. It is also the main field responsible for the growth
of Latin America and Africa as emerging destination regions (the rate of researchers
from this field is especially high for visits to Mexico, Argentina, Chile, Cuba, Egypt and
Morocco). Even though Western Europe continues to be the main destination (60% of
visits), researchers in this discipline make significantly less visits to this region and
North America than the rest of disciplines and significantly more to the above
mentioned emerging destinations, especially over the last decade.
SEJ: Social, Economic and Juridical Sciences (1138 researchers)
The pattern in this field is very similar to that of the Humanities. The discipline is
characterized by a rate growth and a total number of visits per researchers above the
population average, and by a higher proportion of shorter visits (from 1 week to 3
months). Like in the Humanities, visits longer than 12 months are significantly lower
than in the rest of disciplines. As far as destinations are concerned, Latin-America
stands out as being more preferred in this field than in others and North-America as a
less preferred region.
36
FQM: Physics, Chemistry and Mathematics (725 researchers)
This group shows different patterns to the above two. The frequency of short visits is
lower while visits lasting between 12 and 24 months are significantly higher. Western
Europe and North America are highly preferred destinations compared to the above 2
disciplines. Latin-America has a lower rate of visits in this field.
TEP: Production Technologies (725 researchers)
Mobility in this field is significantly lower than in others. Visits per researchers are
therefore significantly lower than the average and most researchers register only 1 visit.
North-America stands out as the most important destination region compared to other
disciplines.
BIO: Biology (441 researchers)
Mobility in this group is significantly higher than the average. The growth tendency of
visits per researcher is however much more stable than in the three first groups which
show high levels of growth over the last decade. The duration of visits in this field is
significantly higher than in others. It is the group with the highest level of visits of 12 to
24 month duration. Western Europe and North America are significantly highly
preferred regions compared to other disciplines. The opposite occurs for Latin-America.
Mobility in this group was higher than the average during the period 1980-1990 and
lower in the decade 2000-2010.
RNM: Natural and Environmental Resources (598 researchers)
The only distinctive features in this discipline are a higher visit frequency level than the
average and a significantly higher rate of longer stays, without attaining the level of the
Biology discipline however.
CTS: Health Science and Technology (451 researchers)
This group shows an overall lower frequency of visits than the other disciplines. This
pattern appears after 2000. Before that year, mobility was higher in this group than in
the rest of disciplines. North America is a significantly more frequent destination region
in Biology.
TIC: Information and Communication Technologies (417 researchers)
The patterns in this discipline are similar to those in CTS: mobility per researcher below
the average and preference for North-America as destination. This group is different
however because it registers mobility levels which are significantly higher than the
average in the last decade, and lower in former periods.
AGR: Agricultural Sciences (400 researchers)
Finally, this discipline stands out for registering lower mobility levels than the average
but longer duration of visits, as well as a relative reduction in visits frequency over the
last decades in comparison to former periods.
37
2.2.6. Summary of main preliminary findings:
• Methodological results:
o The SICA database allows to easily analyse the dynamics of temporary
geographical mobility of the entire population of Andalusian researchers.
o The study of institutional, sectoral or job mobility would however require
much manual coding work due to the way in which this information is
collected.
• Main analytical findings:
o The rate of researchers which declare at least one international research
visit over their careers in the analysed population is 22%, which implies
that 3 quarters of the population has never visited (or never registered in
their CVs) a research centre abroad. The dynamic analysis shows
however a sustained growth tendency since the mid 1980’s, which
attains a maximum average level of 1 visit per researcher every 4 years
lasting between 1 week and 3 months.
o The gender imbalance that appears for the overall population and the
older cohorts disappears in younger cohorts. Although men were more
likely to undertake international visits than women in the past,
significant gender differences disappear for younger research
generations.
o International visits are growing significantly in traditionally less mobile
disciplines like the Humanities and the Social Sciences. The mobility
patterns appear to be different in these disciplines than in others. Visits
in these groups are more frequent and shorter than in disciplines which
are normally considered as being more mobile. This result has important
methodological implications as well as implications for policy making. A
recent European study on obstacles to mobility12 based on survey data
points out that the number of mobile researchers is higher in the life
sciences than in the social and human sciences. The study considers as
mobile researchers those residing abroad when the survey was
conducted. Our results seem to show however that temporary mobility in
the social and human sciences is more frequent and of shorter duration
than in other disciplines, which would imply that it would be more
difficult to find these researchers residing abroad (compared to those
who, for example undertake post-doctoral stays abroad of 1 or 2 years of
duration). A short-term visitor (of up to 3 months) does not register as
resident in the destination country normally.
o The social sciences and humanities present also specific features in
relation to destination countries and regions. The growth of mobility in
these fields over the last decade has come along with the emergence of
new destinations in Latin-America and Africa and the relative decline of
traditional destinations in North-America and Western Europe (although
these remain the main destination regions across disciplines). The
cultural and geographical proximity of the population under study to
12
Commission of the European Communities (2008) “Evidence on the main factors inhibiting mobility
and career development of researchers” by IDEA Consult (coord.), Fraunhover, ISI, NIFUSTEP, PREST,
SPRU and Technopolis. Contract DG-RTD-2005-M-02-01.
38
Latin-America and Northern-Africa are very likely among the
explanatory factors for this tendency.
o Overall, our findings show interesting change patterns that are worth
further exploring for larger populations:
1. The disappearance of gender imbalances in temporary mobility
across disciplinary fields
2. The growth in temporary mobility of researchers in the social
sciences and the humanities
3. The emergence of new destinations for research visits
4. The different mobility profiles defined in terms of duration and
frequency of visits across disciplines.
5. The changes in the international research landscape that the above
patterns may imply.
2.2.6. Information sources and acknowledgements
The information contained in this report has been obtained through interaction with the
persons responsible for the CVN project at the Spanish Foundation for Science and
Technology (FECYT) and through collaboration with the “Consejería de Innovación,
Ciencia y Empresa” of the Andalusian Regional Government.
The URJC EURO-CV team would like to thank José Manuel Báez and Florencio Núñez
from FECYT and Francisco Manuel Solís and Samaly Santa from the “Consejería de
Innovación, Ciencia y Empresa de Empresa.
The brief description of other information systems provided in section 2.2.1. was
obtained through the following websites:
ANECA: www.aneca.es
CAMPUS VIRTUAL: www.ua.es
http://www.ua.es/es/servicios/si/ite/cv/index.html
CANARIAS-CV:
http://www.gobiernodecanarias.org/presidencia/index.jsp?page=oficinacti01.htm
CARTA: http://www.ctt.upv.es/ctt_ingles/tecnological.html
CV UNICAN: http://www.unican.es/
FENIX: http://www.sigmaaie.org/
GINVEST: http://www.um.es/atica/ginvest---gestion-de-la-investigacion
GREC: https://webgrec.ub.edu/
IKERTU: http://www.ikerkuntza.ehu.es/p083-9392/es
NOVATORES: http://usal.novatores.org/html/es/?_language_=es
RIGA: http://riga.xunta.es/cgi-binn/entrada.cgi?nav=s
SIC-MEC/SIC-CINN: www.micinn.es/siccinnciudadano/entrar.do
UCLM Investigación: www.uclm.es
39
2.3. PORTUGAL
Margarida Fontes, André Pirallha, José Assis; DINAMIA.
2.3.1. Assessment of existing CV databases13 (Work Package 1)
There are currently two main databases in Portugal regarding researchers’ electronic
CVs:
• the FCT-SIG (Council for Science and Technology-Information and
Management System) – hereafter called the FCT Information System, and
• the DeGóis Platform
Although this report will be concentrated on discussing the DeGóis Platform, since the
latter will prevail in the medium/long term, replacing the former one, we believe it is
useful to shed some light on the FCT Information System, as it still comprises the bulk
of researchers’ electronic CVs in Portugal.
The FCT Information System (FCT-SIG)
The FCT-SIG is managed by FCT (Council for Science and Technology), which by
Law is required to continuously promote the advancement of scientific and
technological knowledge in Portugal, exploring opportunities that become available in
any scientific or technological domain to attain the highest international standards in the
creation of knowledge and to stimulate their diffusion and contribution to improve
education, health, environment and the quality of life and well being of the general
public.
This mission is mainly accomplished through the financing subsequent to the evaluation
of the merit of proposals presented by institutions, research teams or individuals in
public open calls, and through cooperation agreements and other forms of support in
partnership with universities and other public or private institutions. The results of the
activities of FCT are, in essence, the additional contributions of individuals, research
groups and institutions who have been awarded financing.
It was within FCT’s mission context that the FCT Information System was implemented
by the end of the 90s with the aim at providing researchers and institutions with access
to information relevant to their current relations with FCT. As the researcher’s
participation in submission processes (ex: applications to scholarships, grants for
research projects, research units, publications, etc.) increases, additional options are
made available to in the browser window providing access to elements relevant to the
management of his/her different participations, such as applications results, report
reports, etc. The ultimate goal of this tool was to make interaction between FCT and
system users, essentially the scientific community in Portugal and abroad, more friendly
and transparent.
13
The description of the CV databases relied on information available in the respective webpages:
Plataforma DeGóis http://www.degois.pt/index.jsp?lang=en; FCT Information System -
http://www.fct.mctes.pt/fctsig/; as well as on interviews conducted with a former high-level FCT public
officer, who was involved in the implementation process of the FCT Information System and one
member of the Plataforma DeGóis management team, who was involved in the standards definition and
in the implementation process since its inception.
40
A major feature of this database is its strict confidentiality, as long as there is no free,
open access to researchers’ electronic CVs. Furthermore, it is quite difficult to access it,
bearing in mind that its low level of data standardisation does not allow any meaningful
study regarding S&T indicators. It was also intended to remain active only for a short
period, until the DeGóis Platform was operational. However, difficulties surrounding its
technical definitions and implementation imply that the FCT Information System is still
fully active; it currently covers more than 15,000 CVs.
The CV Webpage gives the researcher access to the tools needed to introduce and
manage his/her résumé. For that purpose, all one needs to do is type his/her public key
before starting filling in the application form. Once introduced in the system, the CV
data will be available to use in all applications and reports electronically submitted
through the FCT website. Table 6 presents the main categories present in this CV.
Table 6: Fields in FCT-SIG Curriculum Vitae
1. Personal data
2. Academic degrees
3. Previous activities
4. Area of scientific activity
5. Present research interest
6. Supervising experience
7. Participation in research projects
8. Prizes and Awards
9. Publications
10. Communications
11. Language skills
The FCT-SIC could provide a valuable repository of data14. However, the system used
to collect and store the CV information does not ensure data standardization and does
not allow the systematic retrieval of the data. In addition, the nature of the data does not
allow the extraction of unidentified datasets, which would be a requirement given the
need for confidentiality. This means that this specific “electronic CV” does not present
the characteristics that make this type of source particularly advantageous.
The limitations of the existing system led FCT to contract the development of a more
sophisticated instrument, leading to DeGois Platform.
14
The FCT /SIG comprises around 15000 CVs.
41
The DeGóis Platform
The DeGóis Platform is an instrument for gathering, supplying and analysing the
intellectual and scientific production of Portuguese researchers. It consists of a portal
having as main features the individual management of the curricular information, the
visualization of national science and technology indicators and the search for curricula
according to content related queries.
The DeGóis Platform is owned by FCT, part of the Portuguese Ministério da Ciência,
Tecnologia e Ensino Superior. Through a protocol established with the Brazilian
Science and Technology Ministry, the GAVEA laboratory from the Department of
Information Systems of the University of Minho, and the STELA research group from
the Federal University of Santa Catarina in Brazil, it provides the basic principles of the
DeGóis Platform and establishes a legal-institutional framework in which the project
will be developed.
DeGóis Platform offers more advanced functionality than FCT own FCTSIG system in
what concerns curricula. Thus, in due time, the DeGóis Platform is expected to replace
the FCT Information System. It can also be used as an instrument for the international
evaluation of Portuguese universities and other research institutions in the country. It is
an (almost) fully public database, apart from some personal data that are not accessible.
Furthermore, it was built with a view to achieve a high level of data standardisation that
would allow for meaningful studies regarding S&T indicators. Below we will analyse in
more detail the main features of the DeGóis Platform.
2.3.2. The DeGóis Platform
Brief Historical Context
The DeGóis project began in 2001, initially developed by GAVEA, following a request
issued by the OCES – Portuguese Observatory for Science and Technology. It gave
continuity to the work started in 1995, in the context of the SICT (Sistema de
Informação sobre Ciência e Tecnologia), aiming at the development of a scientific
platform that could provide for a systematisation of the information concerning
researcher’s individual paths.
In 2001, because of several contacts and some improvements implemented in the
project, OCES found that it was important to invite GAVEA and other foreign
organisations to become partners in the DeGóis project. Contacts were established with
LATTES, in Brazil, through the CNPQ – Conselho Nacional de Desenvolvimento
Científico e Tecnológico (National Council of Technological and Scientific
Development), and they reached an agreement that allowed the use of the Brazilian
LATTES platform after some adaptations to the Portuguese reality (there were not only
significant linguistic differences between Portugal and Brazil but also different CV
presentation cultures). In addition, this agreement required the involvement with the
ScienTI network (see next point), in charge with the system’s maintenance and
continuous technological update, what came to happen in 2003.
This partnership with LATTES and ScienTI was of great importance for the
development of the DeGóis project insofar as it enabled the sharing of technology and
methodology. Nonetheless, at the same time, another European information platform
42
named CRIS – Current Research Information System (see next point), was also
emerging. Although CRIS and DeGóis both had similar objectives (the gathering of
information in order to assist the shaping of S&T management and public policies), they
differed on the data gathering sources: the former gathers its data from research projects
and the latter from researchers’ CVs. However, especially after a meeting between both
working teams in 2005, both platforms share concern for keeping the data exchangeable
and in accordance with several international S&T information standards.
Networks and Partnerships
DeGóis Platform is member of a number of international networks concerned with
development of international CVs and working to ensure the inter-operability between
countries. Of particular importance are:
ScienTI (www.scienti.net) is a public network of sources of information and knowledge
that aims to contribute to the management of scientific, technological and innovation
activities. The network provides a public and cooperative space for stakeholders of
national science, technology and innovation (ST & I) systems and communities of the
member-countries to interact. The network is the expression of international
cooperation among the National Science and Technology Organizations (ONCYTs),
International Science and Technology Cooperation Organizations (OICYTs), Groups of
Research and Development in Information and Knowledge Systems (GDIs) and
Sponsoring Institutions (SIs).
EuroCRIS (www.eurocris.org) is a non-profit association aiming to become the
internationally recognized point of reference for all matters relating to Current Research
Information Systems (CRIS). The association owns a trademark and registered Internet
presence (www.eurocris.org) which also provides the official contact information. The
association unites experts in the field of CRIS and through its strategic partnerships also
all major stakeholders in the European research community, and brings its members
biannual meetings, workshops, an annual seminar and every two years a conference, as
well as on/line discussion forums with peers and publications.
DeGóis Platform is an important partner in both networks working towards the
development of a European platform of CVs.
Current and Future Implementation
It is expected that in the near future the DeGóis Platform will replace the current FCT
Information System, still in use by the Portuguese Council for Science and Technology,
which has shown a number of shortcomings since the moment it was implemented, back
in the 90s.
The DeGóis platform is already an alternative platform for people applying for
fellowships and research projects, and several efforts were made in order to obtain the
universities’ acceptance. Indeed, Portuguese universities as well as other research
and/or scientific institutions can formally become, upon request, a member of the
platform – there are around 40 universities and other research organisations that have
already joined the platform, although their level of participation differs widely. Being a
member of the platform carries a number of potential advantages to these institutions
i.e., they integrate the portal, creating a link to it and they have the possibility of
producing indicators for the institution and for its researchers as well as of
43
automatically obtaining the respective reports. Moreover, any Portuguese researcher,
regardless of working or not for an institutional member of the platform, can create a
DeGóis curriculum.
Nevertheless, the process of replacement has proven to be quite lengthy because there
are several aspects to consider. First, there was some delay caused by the adaptation of
the Brazilian LATTES platform to the Portuguese scenario and because of the OCES
statistical pre-requisites that took longer than expected to insert into the DeGóis
platform. In addition, the filling of DeGóis CV is very time consuming and thus it may
take some time for those scientists who had already filled a CV in the previous system
to move to the new system. Furthermore, many researchers react against the lack of
confidentiality, once their CVs are part of the DeGóis database.
Thus, the introduction of new CVs depends on the initiative of individual scientists, as
well as on the agreements established with some universities with a view to achieve the
migration of data from local CV systems to the Platform. As a result, the number of
CVs registered in the DeGóis database is still low (around 6000 in May 2009) and there
is a strong bias towards some institutions, namely those that joined the Platform in the
experimental stage and whose researchers’ CVs were already introduced in the database
(manually or electronically). Therefore, this database is still far from mirroring the
Portuguese S&T system, both in terms of numbers and in terms of institutional and field
representativeness.
Some Important DeGóis Features
According to the documents published by the platform management it has a number of
features that are described below:
- Evaluation – Monitoring: At the institutional level, the DeGóis platform allows for
an accurate search of specific scientific and technological production indicators.
Scientific field, sectors (socio-economic objectives of each specific field),
keywords, institution and degree of researchers classify this data. According to the
team’s viewpoint, this data can be particularly important for the research
organisations: these indicators can be of use to the universities and other research
organisations for its performance evaluation as well as for human resource
management, supporting career evaluation, and promotion of its own researchers.
- Interface: A DeGóis CV presents a friendly and uniform interface, compatible with
other types of systems of data collection and commercial software.
- Accuracy: The DeGóis platform has several mandatory fields that must be filled
using a pre-determined terminology and, if missing or filled incorrectly, the system
will acknowledge it and will ask for a correction before it allows the CV’s
submission. The DeGóis platform emerged after a thorough analysis of several other
platforms. Therefore, its authors tried to avoid the same early common mistakes and
aimed to improve its flexibility. The structure and type of information provided by
the platform is subject to a high standardization that implies a hidden, but as
uniform as possible, work of categorisation. This means that although most part of
the data is allowed to be shaped at the researchers’ will, those fields regarded as
fundamental to characterise the development in science and technology as well as
those that can provide accurate information about the country’s S&T are
compulsory (in practice they cannot be left vacant).
44
- Openness: Any researcher can access and register himself/herself into the DeGóis
platform. The DeGóis project assumes the complete accessibility of the CV, the
only exception being the identification data (identification card). The DeGóis
working team explained that this principle is not easy to manage since there are
different positions among individual researchers and policy officers regarding the
need to maintain confidentiality. The principle is that all CVs in the DeGóis
database should be publicly accessible. Every CV is reachable through Google and
can be copied and downloaded.
- Comparability: With the objective of identifying the scientific domains of the
researchers’ work, the DeGóis platform allows for the establishment of relations
between the scientific production and the Fields of Science (FOS) established by the
OECD (2006 version). The use of an international standard makes it possible to
compare the DeGóis curricula with other models coming from different scientific
communities.
- Visualisation of Curricula: In the section dedicated to visualising curricula, any
registered user can perform CV searches by name, researcher's institution or region.
Once one has selected the criteria, the system presents the relevant CVs that can be
visualised in an HTML page containing all the information inserted into the CVs.
- Impact Factor: At present, there is no information about the impact factor because
there are difficulties in finding and classifying all the journals within each scientific
field. This work is currently being done.
Curricula Content
The curricula management system (DeGóis curriculum) allows the researcher to insert
his/her personal data, both personal and professional addresses, academic activities,
spoken languages, prizes and awards, research fields as well as all other kinds of
scientific production and detailed information about research projects in which he/she is
currently engaged. It can also include information related to supervisions and jury
participation. The Platform also offers the researcher the possibility to on-line print the
CV and to import and export data.
Briefly speaking, the DeGóis curriculum is structured in three modules:
- General Data, that includes the researcher identification, professional and home
addresses, academic background, professional background, scientific fields, spoken
languages, and prizes and other awards
- Projects, displaying all R&D projects where the researcher currently participates or
has participated before
- Production, including the researcher’s scientific, technical, and artistic/cultural
production (see Table below), as well as his/her supervision activities.
Table 7: Production indicators in the DeGóis platform
Scientific Production
Books and book chapters
Edited or organized books
Edited book chapters
Papers in journals (with scientific refereeing)
Papers in journals (without scientific refereeing)
45
Papers in conference proceedings (with scientific refereeing)
Papers in conference proceedings (without scientific refereeing)
Texts in periodicals or magazines
Texts in newspapers
Other scientific production
Technical Production
Other technical production
Complementary Data
Supervisions
Other production
Participation in academic degrees jury
Participation in evaluation committees
Participation at events
The DeGóis platform enables anyone who accesses it to have online information about
some Science and Technology indicators. This can be obtained either at an aggregate
(the whole universe of CVs entered into the database) or at an individual (researcher)
level:
At an aggregate level, the platform displays seven types of graphics15:
Researchers by gender and age
Researchers by field of work
Researches by academic level
Researches by academic level and field of work
Production by field of knowledge
Production by field of knowledge and five-year period
Production by type
At an individual level, a number of “Production Indicators” are automatically generated
by the system and are placed at the bottom of each individual CV. These indicators act
as a summary of a researcher’s scientific and/or technological production. Below there
is an example of such indicators:
Scientific production
Technical production
Artistic and cultural production
Artistic and cultural production description
Supervisions
Participation in academic juries and other academic events
Participation in evaluation committees
Participation in other juries
Participation in events and scientific meetings
15
See webpage at: http://www.degois.pt/index.jsp?id=2
46
2.3.3. Exploratory studies (Work Package 2)
Objective of the analysis
The objective of the exploratory analysis was to test the possibility of using data from
Electronic CV databases to build meaningful mobility indicators and of using them to
understand scientific mobility dynamics, as well as to address the impact of mobility
upon career development and scientific production.
Our research started from the assumption that, while electronic CVs permit to overcome
some of the methodological shortcomings frequently associated with traditional CVs –
lack of standardisation, absence of key information, codification difficulties – they still
have some limitations as sources of mobility-related data. In fact, the CV basically
documents career evolution and its outputs and, therefore, it enables the collection of
data on mobility along the scientist career by indicating the location (geographical and
institutional) where main career events took place. However, events that involve
mobility differ in importance, from a career development viewpoint, and thus scientists
may choose not to report all of them in a CV. This means that short-term mobility is
less likely to be reported, unless the events it is associated with are perceived as
particularly relevant, indicating professional or learning status or productivity levels.
Mobility events are especially vulnerable to the purpose of the CV, which influences the
events reported and the degree of detail supplied. Thus, in general the CV is particularly
useful when it comes to draw a general picture of mobility trajectories along the career.
However, in order to use it adequately, it is necessary to take in consideration its
limitations and to assess which is the data that can be really be obtained from CVs and
which needs to be obtained from complementary sources, as well as how to capture all
this information.
In order to conduct a first exploration of the usefulness of electronic CV based data to
build mobility indicators we asked FCT authorisation to use data from the DeGóis CV
database, which was granted provided that confidentiality was maintained. The data was
supplied by the DeGóis management team at Gávea (Universidade do Minho) and
consisted of unidentified datasets for all PhD holders present in the database by
November 2008. These datasets contained a sub-set of variables that were selected from
the list of variables supplied by the DeGois team according to two criteria: a) was
judged to be relevant for an analysis of mobility and its impacts; b) was acknowledged
by DeGois team to display a reasonable degree of completeness16.
The analysis followed three steps. We started by conducting a first assessment of the
data supplied, which led us to conclude that it suffered from several problems – limited
representativeness; problems at the level of data completeness, namely in some key
variables (e.g. publications); some “teething problems” in what concerns the level of
standardization and accuracy of the data - that are discussed in more detail below. While
these problems strongly constrained the empirical research that could be conducted at
this stage, we considered that most of them are related with the early stage of
development of the Platform and will be solved when DeGóis CV filling effectively
becomes mandatory to all researchers who wish to receive public funding from FCT
and when some database structure/organisation issues are addressed. Once these
difficulties are overcome, it is expected that the data obtained from DeGóis Platform
16
The availability and support of GAVEA researchers and particularly of Doutor Leonel Santos, who
contributed to the definition of the relevant variables, is gratefully acknowledged.
47
will allow for the construction of indicators that will ultimately enable us to: a) study
mobility trajectories and identify mobility patterns associated with formal training and
professional activities (through data categorisation and analysis of longitudinal career
data); b) evaluate the impacts of mobility on career evolution and on scientific
productivity
Having this in mind we concluded that the analysis conducted in the context of this
project could only have a methodological goal: to assess the usefulness of the data
available in the deGois database to build indicators that enable us to measure mobility
and its impacts. Therefore we adopted the following strategy:
1) to use the documentary information on the data that could be obtained from
DeGóis to conduct a first definition of the type of mobility and mobility impact
indicators that can be built on the basis of electronic CV data;
2) to conduct a preliminary empirical test of some of them, on the basis of the
(limited) data available at this stage.
Given the extensive difficulties confronted when dealing with the data supplied, which
led to a great delay in the development of the project (see below), these activities will
only be completed after the writing up of this report and will be presented later, in two
papers scheduled for scientific conferences17. Thus we will mainly focus on the
characteristics of the DeGois CVs data and on the presentation of the planned empirical
analysis.
Preliminary assessment of the data
The data requested by the EURO-CV team was included in the following categories:
training and professional activities, from which it was expected to obtain core mobility
data; demographic information; and data that could be used to measure mobility
impacts: scientific and technical production, research projects and supervisions (Figure
11). It was decided not to ask for other complementary data that could be used to
measure short-term mobility, since it was immediately assumed by DeGois team to be
very infrequently filled.
17
Pirralha, A., M. Fontes and J. Assis (2009) Assessing Scientific Mobility Dynamics and Impact: the
case of mobility during the PhD, 9th Conference of European Sociological Association, Lisbon, 2-5
September 2009; Fontes, M. and A. Pirralha (2009) Assessing Scientific Mobility Dynamics and Impact:
drawing on the potential of electronic CV databases, 2009 Annual Meeting of the Society for Social
Studies of Science, Washington, DC, October 28 to November 1.
48
Figure 11: DeGóis data categories
Data on scientific production (publications) ended-up not being supplied since it was
very incomplete (only about 1/3 of the sample had included at least one publication).
Table 8 lists in detail the items requested, displaying in red the ones that ended up not
being supplied.
Table 8: DeGois data requested and supplied
Data Categorie Itens
Demographic Sex; Date of Birth; Professional Address; Nationality
Information
Academic Background Degree Title; Institution; PhD course; year of conclusion; ISCED Area; FOS/OCDE
(PhD and Post PhD) Area
Professional Activity Institution; Type of professional career; Type of professional category; % of
dedicated time; Start Date; End Date
Scientific Outputs Journal articles - authors, authors institution, year, publication country, part of the 5
most relevant works?, Journal, ISSN, peer reviewed, FOS/OCDE Area
Books - Authors, authors institution, year, country, part of the 5 most relevant
works?, ISBN, FOS/OCDE Area
Book Chapters - Authors, authors institution, year, country, part of the 5 most
relevant works?, ISBN, FOS/OCDE Area
Edited Books - Authors, authors institution, year, country, part of the 5 most relevant
works?, ISBN, FOS/OCDE Area
Patents - Country, type, code, title, deposit date, exam date, concession date
Research Activity Research Projects - institution, position in the project, national or international, year
began, year finished, collaborations, partners, FOS/OCDE Area
Supervisions – Year; Country; Institution; Course; ISCED Area; Concluded ?;
FOS/OCDE Area
When the EURO-CV team conducted a first assessment of the data supplied, a number
of unexpected problems were identified. These problems are mainly related to a) the
structure, and b) the organisation, of the database and are described below.
49
a) Structural Problems:
Level of data standardisation and (in)accuracy
One of the main advantages of an electronic CV system is the level of data
standardisation implying, on the one hand, the existence of mandatory fields and, on the
other hand, the utilisation of pre-defined terminologies through pre-established lists.
However, once we received the data from the DeGóis management team, we were
confronted with several inconsistencies, namely the use of terminologies not included in
the lists (standardisation) as well as the lack of data in fields that are labelled as of
mandatory [data (in)accuracy].
Level of completeness
The second structural problem relates to the fact that there is no way to guarantee an
exhaustive filling of the several fields in the DeGóis CV by the researcher. Thus, the
degree of fields’ coverage is still rather limited, namely those fields related to Projects,
Patents, and Publications. Indeed, a large number of scientists do not fill – or do so only
in a very limited way – these fields; and when they do fill them, one is confronted with
a large number of “missing” data i.e., fields that are not mandatory are usually not filled
in, which leads to problems regarding the completeness of the CVs in the database.
Composition of the population
The third structural problem deals with the composition of the population part of the
DeGóis database. As everyone can fill in his/her own CV, one ends up with a
heterogeneous population, including researchers and non-researchers. It is therefore
important to clearly define who is considered relevant when producing S&T indicators
on the Portuguese scientific system. Indeed, one might be confronted with these three
situations: Portuguese scientists working abroad, foreign scientists working in Portugal,
and foreign researchers with no links to Portugal who only want to fill in a DeGóis CV.
We believe that three fundamental elements must be taken into account in order to
locate and individualise the population relevant for the construction of indicators:
Professional Activity, Nationality, and Country where is located the institution in which
the researcher is currently working. For the latter two elements, there is a complete lack
of data, because they are not part of the DeGóis mandatory fields, so it is not possible to
analyse these situations.
b) Organisational Problems:
Regarding the data format provided by the DeGóis Platform, the key problem relates to
the assignment of data for each individual CV. For example, when we tried to study the
trajectory of a scientist and identify his/her mobility patterns during the career, we were
confronted with a number of bottlenecks in establishing a sequence of activities and the
researcher’s current situation. It seems therefore important to rethink the ways of
organising the data and/or treating the information.
Following this assessment we produced a report to the DeGóis management team,
describing the problems and the main bottlenecks identified and making some
corrective suggestions and also some recommendations concerning an eventual future
data supply.
50
Building mobility indicators based on the DeGóis (future) data
As was already mentioned, the DeGóis CV database potentially offers a vast amount of
information concerning career, education and scientific outputs. The data available
within these categories, alone or in various combinations, can be used to build a variety
of indicators about mobility and its effects, when the Platform will be fully operative.
Since our objective is to assess the type of mobility indicators that can be obtained from
this specific electronic CV database, the definition of the focus of our analysis had to
take in consideration the already mentioned structural limitations of the CV as source of
information on mobility – focus on professional activities and key training events and
lower reliability regarding short-term mobility events – as well as the specific
characteristics of the DeGóis Platform, which targets particularly (although not
exclusively) scientists who are in the academic/research career and are located in
Portuguese organizations. Therefore we focused on medium to long term international
mobility for professional or advanced training purposes, by scientists who are active in
academic/research careers.
One frequent shortcoming in mobility studies is the absence of a clear definition of the
concept of mobility used. Given the complex and multidimensional nature of the
mobility phenomenon, a clear definition of the dimensions considered is critical in order
to avoid methodological ambiguity.
In this analysis we define mobility as the conduction of professional or advanced
training activities in organizations located in a country that is different from the country
of origin, by scientists (doctorate holders18) who are active in academic/research
careers.
In our definition of scientific mobility, we are effectively taking into consideration a
number of dimensions that we judge relevant to delimit and characterize the
phenomenon we are addressing: time, space, sector of activity. These dimensions can
be used to build a framework that supports a clear delimitation of the concept of
mobility used by the researcher, as well as the identification of patterns in researchers’
mobility:
a) the temporal dimension corresponds to the duration of mobility events: short,
medium or long term mobility – we chose to focus exclusively on medium to
long term mobility.
b) the spatial dimension that refers to mobility between countries or regions and
between organizations – we chose to focus exclusively in mobility between
countries (mobility between organizations being only considered when it
involves change of country).
c) the sectoral dimension that refers to mobility between sectors of activity (e.g.
research vs. non-research; public research organizations vs. industry) and
between scientific areas – we chose to focus exclusively on researchers in public
research organizations, thus excluding mobility between sectors and addressing
only mobility between scientific areas (including all scientific fields).
18
We focus on doctorate holders, given their key role in knowledge production and dissemination Auriol
et als., 2007).
51
Table 9: Dimensions of Mobility
Indicators Mobility
Definition
(DeGois Platform) Dimensions
Medium/ Long - More than a year stay in a Duration of Professional or
term mobility different organization for Training Activity Temporal
professional or advanced
training activities
International Different countries in which Location of Professional or
Mobility the researcher has worked or Training Activity: Spatial
conducted advanced training Country/Organization
activities
Field Mobility Change in scientific area Field of knowledge
FOS/OECD classification Sectoral
This table presents the characterizing dimensions of mobility we defined in our case and the type of data
(available in DeGois Platform) that can be used for the respective mobility analysis.
The indicators resulting from the application of this framework can be used for different
purposes. Firstly they can translate broad tendencies: by allowing the collection of
aggregated data for host country, duration of stay, scientific area, time periods when
mobility took place, they permit to globally characterize mobility flows. Combining
them with individual characteristics of researchers (e.g. age, gender, nationality) they
enable the further identification of patterns concerning the structure and dynamics of
scientific mobility, for the whole population of for sub-sets of that population. They can
also be associated with specific stages along the scientists’ career trajectory with a view
to obtain a more in-depth understanding of mobility dynamics. Secondly, these
indicators can be related with additional dimensions, like career status or scientific
productivity, in order to explore the eventual impacts of mobility. For instance, it is
possible to conduct a comparison, along several indicators, of researchers who engaged
in international mobility with those who did not, taking into consideration the patterns
previously identified, e.g. regarding mobility differences in terms of scientific fields,
time periods / age cohorts, host countries, gender, etc. The results of this type of
analyses can be a useful contribution for an assessment of mobility policies.
2.3.4 Preliminary empirical analysis
Data availability and limitations
As a result of the preliminary assessment conducted on DeGóis CV data, we have
concluded that only the following sets of data could be fully used at this stage:
• Demographic data: age, gender
• Data on the PhD: location, date, area
• Data on current professional situation
While data on full professional trajectories was also supplied, the exercise of extracting
the current situation from this dataset has uncovered several problems (both related with
format in which the data was supplied and with actual data contents), that created great
difficulties in extracting the full set of activities for each case. The attempt to overcome
52
these difficulties would cause an even more considerable delay in obtaining a dataset for
the analysis, preventing us to produce any empirical results within the EURO-CV
deadlines. For this reason and also the given the doubts regarding the reliability of the
data to be obtained, it was decided to leave this dataset for a subsequent stage. The
absence of this data precluded an analysis of mobility trajectories (except for the PhD)
and also limited considerably an analysis of impacts of mobility on career development.
Considering the above, it was decided to conduct a limited experiment on the uses of
trajectory data on the basis of a smaller sample, which included only the case of
scientists who had done a PhD abroad and for which we had complete data on all the
other variables. For this group (103 cases) we have extracted manually, from the
“professional trajectories” dataset, additional data on the professional situation
immediately before and immediately after the PhD. This data, once standardised and
codified following the same criteria as the one used for the current situation, will enable
us to have information on four moments:
• Last job with a date previous to PhD award date (institution, career/position,
start date and end date)
• PhD (institution, award date)
• First job after PhD award date (institution, career/position, start date and end
date)
• Current job (institution, career/position, start date)
This may provide the basis for some very exploratory analysis of trajectories for this
group, which we expect to conduct until the end of September. In addition, we will try
(conditional on availability of human resources) to define an equivalent sample (namely
in terms of age, gender, field) of scientists who did their PhD in Portugal and equally
extract manually the same set of data. This would enable us to conduct a more detailed
analysis of trajectories and assess eventual differences in career evolution.
Characteristics of the global dataset
The data was supplied in two different batches whose compositions did not completely
match. Therefore the dataset on training (PhD degree) amounted to 825 cases (not
excluding cases with missing values in some key variables), while both the dataset in
demographics and the one on professional activities19 amounted to 580 cases (not
excluding missing values). Matching the two datasets produced a final sample of 520
cases. After accounting for missing values in key variables (data for some fields was
particularly incomplete) and for inconsistencies, we ended-up with 500 matching cases.
It is relevant to point out that significant numbers of missing values were found in
variables that were described as compulsory and that, in principle, could not be left
unfilled. If we consider only the cases where there are no missing values in all the
variables, we are reduced to 367 matching cases20.
Some other aspects deserve to be mentioned regarding this sample sample: a) 4
universities (which include the two Plataforma deGois pioneers – ISCTE and
19
This data was also supplied in first batch, but incomplete (without start and end dates). Complete data
was only obtained in 2nd batch and thus this dataset was reduced to 580 cases.
20
However, the number of cases effectively depends on the type of analysis being conducted.
53
Universidade do Minho) represent about 52% of the sample21 (see graph 1 in the
appendix); b) there is a strong bias towards scientists with career positions: “out of
career” scientists (i.e. those with grants or short term contracts) account for less 10% of
the sample (from data on career/position); c) there is a strong bias towards scientists
currently in Portugal; only 4 exceptions.
Table 10 presents the basic variables that were retained for the final analysis and
indicates the main problems associated with some of them. Table 11 presents the
additional data obtained manually for the smaller dataset described above.
A more detailed characterisation of the sample will be conducted later. At this stage, we
present some information in the Appendix, including generic data (distribution by
institution, scientific field, age) data on international mobility (countries, scientific
fields, year) and data on current position (start year, type of career and career position).
Table 10: Variables from Plataforma DeGóis used in the analysis
Dimensions Variables
Training: PhD Year conclusion PhD 1
(1st batch) Institution PhD
Country PhD 2
Area PhD (FOS classification) 3
Training: Post-doctoral activities 4 Year conclusion Post-doc 1
(1st batch) Institution Post-doc
Country PhD 2
Area PhD (FOS classification)3
Demographics data Date of birth
(2nd batch) Gender 1
Current Location (geographical) 5
Current professional situation Current Institution 6
(2nd batch) Country of current institution 2
Current career type 7,8
Current position in career 7
Year started current position
Legend
1 – Have missing data in significant numbers
2 – Codified by us
3 – In principle, the areas had been classified using both FOS (Fields of Science) and ISCED, but the
actual data contained in the respective fields did not always match ISCED and FOS classifications:
several fields had been added to it. Combining the data from the two fields we attempted to reach a final
codification, using the 7 main FOS categories: Exact Sciences & Natural Sciences (disgregated in the
Portuguese version of FOS); Engineering and technology; Medical and Health sciences; Agricultural
sciences; Social sciences; Humanities.
4 – Only 39 cases (before excluding missing values). The respective variables were added but ended-up
being discarded given their limited usefulness.
21[4]
The sample includes 62 institutions, the vast majority of which (including some major universities)
contribute with less than 10 cases to the sample.
54
5 – Since the vast majority of the researchers are currently located in Portugal, the geographical location
was codified by municipalities and an additional category “abroad” was included.
6 – Two potential sources: a) “institutional address” – which can conceal cases of mobility (keep address
but is currently elsewhere) and has no additional data associated (position / year); b) Institution of current
position – extracted from database on trajectories using as criterion “position that started in more recent
year/month that has no end date”. Problems: cases of more than one position with no end date where last
one is not necessarily the main position; cases missing end dates. Institutions from both sources do not
always coincide. Those cases were cross-checked manually in the database, at the light of previous
trajectory and sometimes required us to make a decision concerning the choice of final institution.
7 – See above comment on problems in identifying the “current position”: when more than one position
without end date a decision had to be made concerning the one to choose, since the most recent start date
might not always be the more adequate. Other criteria were: level of stability (e.g. visiting professor vs.
tenured position) or higher career position.
Table 11: Variables from Plataforma deGois for the experimental analysis of
trajectories
Dimensions Variables
Professional situation before PhD award Institution
Country of institution
Career type
Position in career
Year started position
Professional situation after PhD award Institution
Country of institution
Career type
Position in career
Year started position
Note:
This data was obtained by extracting manually the respective fields from the whole “professional
activity” dataset, searching for the activities that started in the date closer to (but before) the PhD award
date, for the “situation before”, and the activities that started in the first year after or in the same year of
the PhD award date, for the “situation after”. When more than one activity existed in the same date, a
decision was made, following the same criteria described for the “current situation”. Notice that in some
cases there was no end date, so the situation after the PhD coincided with the current situation. There
were also a few cases whose “situation after” started before the PhD, thus the “situation before” coincided
with it (and sometimes also with the “current situation”). Finally in some cases there was no information
about the situation before the PhD and also, in a few cases (namely of scientists already in middle or top
career positions), there was only one register, which was corresponded to the current situation. This
confirms the problems of incompleteness of data.
Regarding career and career position, the data was organized as follows. We defined 3
main careers in the scientific profession: researcher, university professor, polytechnics
professor. In the case of position previous to the PhD, there was also the possibility of
being outside the scientific profession and therefore we included one “other” category
for this purpose. Career positions were more complex to organise, given the
proliferation of levels and the presence of parallel levels in the different careers.
Additionally we also had to deal the presence of a number of researchers who are
outside the existing careers, being hired under a variety of situations (grantees, short-
term contracts in the context of projects; post-doctoral scholarships, etc), which were
not contemplated in the Plataforma deGois structure and who used a wide variety of
55
designations to define their situation. Therefore, we first defined levels that were
equivalent between three careers and then aggregated them in 5 main categories:
- Pre-PhD positions
- 3 levels for Post PhD positions: top, middle and start
- “Out of career” post PhD
The work conducted to organise the data supplied in a format that could be used by a
statistical package such as SPSS, to match the datasets supplied separately (and
particularly the different datasets supplied in the two batches) and to extract the
information about current situation from the dataset on professional activities, required
computing skills that went much beyond the competences of the EURO-CV team. For
this purpose we resorted to the support of our colleague Joaquim Duque, from INETI’s
Departamento de Modelação e Simulação, whose help is gratefully acknowledged.
Nevertheless, given the characteristics of the data, the automatic extraction and
organisation of the data had to be complemented by extensive manual checking and
correction. In addition, several presumably “standardised variables” were not accurate
and had to be recodified by hand. While this exercise is possible (although very time
consuming) in this still relatively small dataset, it will become much more complex as
the database becomes larger. In addition, given the nature of the data on trajectories, it
was judged more effective and reliable to recover it manually. It would thus be
important to correct some of the problems identified in the database, in order to fully
benefit from the advantages of electronic CVs, concerning standardisation and
completeness of the data.
Analyses to be conducted with the CV data available
a) Mobility for training:
Regarding the whole dataset, the data available permits to identify some patterns at this
level:
- Proportion PhDs abroad & Differences according to age cohorts / time periods,
gender, scientific field.
- Preferred countries for PhD & Differences according to age cohorts / time
periods, gender, scientific field.
- In addition: some analysis of national context (preferred national universities for
PhD and differences by scientific field)
b) Influence of training abroad in (current) professional situation
Regarding the whole dataset, since at the moment we only have the current position it is
impossible to make an analysis of impact of international mobility upon career
progression, because we do not know which was their first position after the PhD.
However we can assess:
- whether scientists who did PhD abroad are currently in Portugal or abroad
(problem: strong bias towards scientists in Portugal in the data)
- how many years passed since completion of PhD and which is career position,
namely comparing different age cohorts relative to it; whether there are
differences between those with /without international mobility in PhD
56
- how many (from recent cohorts: e.g. after 1995 or 2000) are not yet in career;
whether there are differences between those with /without international mobility
in the PhD.
- differences by scientific fields
- In addition: can assess mobility within country borders: change of institution;
change of location (region) as compared with PhD.
c) Exploratory test of trajectory data
Regarding the smaller data set, its purpose is basically to explore the possibilities of
using data on trajectories. Since it is composed exclusively of scientists who did a PhD
abroad it will enable us to conduct such exploration on this specific group, although it
does not enable us to address the impact of mobility on career evolution. Considering
the nature of the data it will be possible to test the use of this data to address:
- early trajectories for different age cohorts / time periods, gender, scientific field
- career evolution (early vs. current) and differences
If we manage to also extract the same set of data on an equivalent sample of scientists
who did the PhD in Portugal, it will be possible to address:
- early trajectories for the mixed set: regularities and differences between them.
- impact of conducting the PhD abroad in career evolution and eventual
differences between fields
Characterisation of generic sample
Generic characterisation
Figure 12: Institutional bias: Universities and other research organisations in the
sample (N=500)
57
Figure 13: Distribution by scientific fields (N=825)
13% 3%
21% Ciências agrárias
Ciências da engenharia e tecnologias
Ciências exactas
Ciências médicas e da saúde
Ciências naturais
34% 13% Ciências sociais
Humanidades
5%
11%
Figure 14: Age distribution of sample (N=493)
Count of IDADE
35
30
25
20
15
10
5
0
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
IDADE
58
Figure 15: Year completed PhD (N=493)
Total
Count of ANO_CONCLUSÃO
60
50
40
30
20
10
0
1965
1968
1969
1970
1971
1973
1975
1976
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
ANO_CONCLUSÃO
Mobility during PhD (N=825)
International mobility for the PhD = 226 cases (23.4%)
Figure 16: International mobility for the PhD: main countries
59
Figure 17: International mobility for the PhD: scientific fields (N=226)
Current position
Figure 18: Date started current position (N=500)
Count of ANO_INICIO
70
60
50
40
30
20
10
0
1976
1977
1979
1981
1982
1984
1985
1986
1987
1988
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
ANO_INICIO
Figure 19: Distribution by career (N=500)
15%
University professor
8%
Polytechnics professor
Researcher
77%
(Note: “out of career” people are included in the “researcher” category)
60
Figure 20: Career position (N=496)
60,0%
50,0% 52,0%
40,0%
30,0%
20,0%
18,8%
10,0% 11,7%
8,7% 8,9%
0,0%
Out of career Pre-PhD Early stage Mid-career Top career
3. Ad-hoc CV databases: the UK experience
Aldo Geuna and Ana Fernández-Zubieta; SPRU University of Sussex.
This report summarises the activity done in the UK case study. The report is structured
in two main sections. Next section indicates the availability, characteristics and
accessibility of electronic CV databases and registers in the UK. The following section
explains methodological aspects of CV collection, coding and analysis in the country.
3.1. Availability, characteristics and accessibility of electronic CV databases and
registers
The decentralisation22 and diversification of educational and research services make it
difficult to find a broad, electronically structured researchers’ CVs’ public database in
the UK. A large variety of organisations require researchers’ CVs in the selection
processes for education and research funds, employment or research awards. The lack
of a single national public research institution redirects attention to the main
organisations that fund research and employ researchers. The main organisations that
fund research and/or employ researchers are reported in order to fulfil the WP1 of the
euro-CV project (‘assessment of existing selected CV databases and data collection’). In
any case, the requirements of the Data Protection Act 1998 limit the access to the CV
records held by these institutions.
22
The country does not have a central register of UK academic staff. The Higher Education Statistics
Agency (HESA) provides data of academic and non-academic contracted staff of Higher Education
Institution (HEI) from 1994/95 to 2002/03. The requested fields are: activity, age, cost, centre, disability,
ethnicity, full-time equivalence (FTE), gender, grade, highest qualification held, location of institution,
mode of employment, nationality, subject of higher qualification, source of salary and terms of
employment.
61
The main organisations that require researchers’ CVs in selection processes are listed
below. A report of the general stored format, uploading culture and degree of
standardisation of researchers’ CVs follows.
3.1.1. Research funding organisation and requirement processes
Research
The UK has a large variety of organisations that fund research (public, non-profit and
private companies) (See Table 1). The main sources of public funding are the research
councils, several government departments and the higher education funding councils.
Other bodies funding research are charities (e.g. Welcome trust, Leverhulme trust) and
learned professional societies. Universities and research institutes also fund research23.
The range of their scope and selection processes for funding allocation varies
drastically. No general procedure for the requirements, availability and storage of
researchers’ CVs is thus found.
A summary of the Research Councils policy of allocation of funds procedure is reported
as a guideline to accepted practice. The demand and availability of researchers’ CVs
required within selection processes is reported.
The Research Councils provide for research, training, knowledge transfer and public
engagement. The main role of the research councils is to fund research24. The Research
Councils are structured according disciplines. The outputs from this research are
publicly available for researchers and other potential users in business, government and
public sectors. However, as indicated above, the Data Protection Act 1988 prevents
access to the personal data of the applicants.
Public availability of research outcomes and policy engagement of the Research
Councils does not include access to complete CVs records of researchers. However,
some of this data (i.e. name, organisation, department and e-mail address) may be
accessible for reporting purposes.
Applications for funds should be submitted to the relevant Research Council through
the common “Councils Joint Electronic Submission System” (JeS). For each applicant
and named research staff a CV in electronic format must be included. The general
guideline provided by the Councils for the CVs’ structure suggests including
information on: contact details, qualifications, academic and professional posts held
since graduation, a list of the most relevant and recent publications and a record of
research funded by the ESRC and other bodies. The CV should not exceed two A4
sides. The limited extension and the non-provision of a standard format determine the
level of standardisation of researchers’ CVs.
Research Councils: Non-public availability and medium standardisation of researchers’
CVs records is found.
23
International sources of funding are not considered.
24
The yearly investment is around £1.3 billion in research in UK universities and £500 million in their
own Research Institutes, and £300 million in access to international facilities for UK researchers.
62
Other organisations and selections processes
Universities, firms and other research related bodies (e.g. Charities) require researchers’
CVs within selection processes. As a general rule, they operate under the Data
Protection Act 198825. Although some exceptions may occur, in general terms, the only
source for researchers’ access to CVs is through personal contact with researchers.
Other research bodies: access and standardisation varies according to the organisation.
Generally, non-public availability and low standardisation of researchers’ CVs records
is found.
25
Previous contacts with the Research Councils did not succeed in negotiating a special agreement.
Contacts with Royal Institutions were also unsuccessful.
63
Table 12: Main research funding organizations
MAIN RESEARCH FUNDING ORGANISATIONS*
• Research Councils
o Arts & Humanities Research Council (AHRC)
o Biotechnology & Biological Sciences Research Council (BBSRC)
o Engineering & Physical Sciences Research Council (EPSRC)
o Economic & Social Research Council (ESRC)
o Medical Research Council (MRC)
o Natural Environment Research Council (NERC)
o Science and Technology Facilities Council (STFC)
• Government departments
o Department for Business, Enterprise and Regulatory Reform (DBERR)
o Department of Health (DoH)
o Department for Environment
o Food and Rural Affairs (DEFRA)
o QinetiQ and dstl (formerly Defence Evaluation and Research Agency-
DERA)
o Department for Innovation, Universities and Skills (DIUS)
o Department for Transport (DfT)
o Department for Employment and Learning, Northern Ireland (DEL or
DELNI)
• Higher Education funding councils**
o Higher Education Funding Council for England – HEFCE
o Scottish Funding Council – SFC
o Higher Education Funding Council for Wales - HEFCW
• Charities
• Learned professional societies
o The Royal Society
o The Royal Society of Edinburgh (RSE)
o The Royal Academy of Engineering
o The Engineering Council (EC)
o The Engineering and Technology Board
o The Royal Institution of Great Britain (RI)
o Foundation for Science and Technology.
• Universities and research institutes
* This list is illustrative. It has not been applied a consistent criterion for its selection. International
sources are not included
** Through DIUS, the councils allocate funding for university basic research in England, Scotland and
Wales. DEL provides funding for research in Northern Ireland.
Relevant websites for location research funding and employment bodies:
• http://www.rcuk.ac.uk/default.htm
• http://www.hero.ac.uk/uk/home/index.cfm
• http://www.researchresearch.com
• http://www.britishcouncil.org/science-uk.htm
• http://www.timeshighereducation.co.uk
• http://www.wellcome.ac.uk
64
• http://www.leverhulme.ac.uk
• http://www.jobs.ac.uk/
3.1.2. Storage format and policy purposes
The storage of researchers’ CVs varies according to the institution. In general terms, an
electronic (pdf.) and paper version is required. When an electronic submission system is
used, an attached electronic (pdf.) version is required. Thus, non-electronically filled
researcher’ CVs records is required (see Research Council procedure above).
The Research Councils and other research bodies play a role in helping to convey
government objectives for science, technology and innovation. They provide
submissions (e.g. evaluations, reports and recommendations) to consultations and
Parliamentary inquiries. In the case of the Research Councils, the use of researchers’
CV records for the elaboration of these submissions is limited to restricted data (i.e.
name, organisation, department and e-mail address) (see above)
3.1.3. Uploading culture and level of standardisation
Researchers in the UK usually upload their CVs onto institutional websites. However,
this practice varies according disciplines. Natural scientists and engineers tend to upload
their CVs records more often than social scientists. The information posted refers
mainly to publications and research lines. In general terms, data on previous positions is
missing. Consequently, job mobility records are effectively not available through
websites.
The large variety of organisations entitled to research and their different practices
makes researchers’ CVs formats very inhomogeneous. The main bodies do not provide
a standardised CV format in the requirements for research funds. As a result, the level
of standardisation of researchers’ CVs in the country is low. The main problem is the
lack of standardisation makes it difficult to complete the information for variables that
could be relevant for analysing mobility. A thorough consideration of the minimum and
most valuable information is crucial (see coding process).
SUMMARY:
Non-existent electronic structured CVs public database is found
Low level of standardisation on researchers’ CVs low is found
Low level of accessibility because of the Data Protection Act 1998 is found
3.2. Methodological aspects of CV collection, coding and analysis in the UK
The absence of a national electronic structured CVs public database in the UK increases
the difficulty of collecting and coding data from CVs. The accessibility of the CVs
collected by important research organisation (e.g. Research councils) is limited by the
Data Protection Act 1998, which makes the researchers' CVs–collection process more
difficult. The codification process is complicated by the low level of standardisation of
65
researchers´ CVs in this country. Therefore, lack of availability and standardisation of
CVs increases the shortcomings of using them for data collection purposes.
The main shortcomings26 of coding data through CVs come from:
• Semi-standardised formatting – CVs contain different information.
• Missing information. It can not be assumed that the lack of information
regarding certain research activities (e.g. patent activity) indicates non-activity.
• Varying length and truncated CVs – e.g. some CVs only refer to a period of five
recent years. The use of truncated researchers' CVs is a common practice in the
UK. Research Councils request recent or selected publications in their funding
and evaluation processes (e.g. Research Assessment Exercise).
The collection methodology process is explained below. Other sources used to complete
CV information, coding methodology and characteristics of the sample follow. The
description of the analysis includes research questions, types of analysis, a summary of
the results and a report of the main significant methodological problems encountered.
The section ends with a list of publications.
3.2.1. Collection methodology
The sample was drawn from the Engineering and Physical Sciences Research Council
(EPSRC). As mentioned previously, researchers' CV records can be accessible for
reporting purposes.27 This access is limited to restricted data: name, organisation,
department and e-mail address.28 The data includes all the principal investigators that
were awarded EPSRC funding for a project from 1995 to 2003 in 10 scientific fields.
The researchers' CVs were requested by e-mail with an additional questionnaire on their
patenting activities. This questionnaire was added due to the fact that patent information
is generally missing in CVs.29 The request of CVs was limited to a 4 disciplines:
chemistry, physics, computer science, and mechanical, aeronautical and manufacturing
engineering. The aim of the focus on these four disciplines is to checki differences in
researcher mobility across the basic and transfer sciences.30
The CVs were used to gather information on changes in job positions and to clean-up
information on publication and patent records – which were obtained through the ISI
Web of Science database and the European Patent Office (EPO) (see Figure 1). This
method of collecting data using CVs, in addition to the ISI and EPO databases,
improves accuracy. Mistakes arising from similar names and initials, along with
changes in researchers’ institution affiliations, are avoided. Collecting data from CVs
26
Dietz, J 2004. Scientists and Engineers in Academic Research Centers – An Examination of Career
Patterns and Productivity. PhD Dissertation. School of Public Policy. Georgia Institute of Technology.
February.
27
The report aimed at providing a better understanding of research collaborations between universities
and industry D’Este, Pablo, L Nesta and P Patel 2005. Analysis of university-industry research
collaboration in the UK, SPRU Report. Falmer: SPRU.
28
Previous contacts with the Research Councils did not succeed in negotiating a special agreement.
Contacts with Royal Institutions ended up were also unsuccessful.
29
Crespi, G., D'Este, P., Fontana, R. and Geuna, A. 2009. The impact of academic patenting on university
research and its transfer. ICER, Working Paper.
30
This distinction follows previous works (e.g. Geuna, Aldo and Lionel Nesta 2006. University patenting
and its effects on academic research: The emerging European evidence. Research Policy, 35, 790-807.)
66
allows job–transition information – as well as reliable information on publication and
patenting activities – to be gathered.
The collection of CVs via e-mail reduces the sample size and introduces biases in
response rates. The lack of standardisation of the CVs made the manual coding of the
information on CVs unavoidable. Manual coding is very time-consuming making it
difficult to work with big sample sizes. The higher response rate of higher academic
positions and basic sciences introduces job position and disciplinary biases. Professors
and chemists and physicists31 were more willing to send their CVs.32 A complementary
web search was done in order to avoid biases and complete missing information.
An attempt to develop a software programme was done in collaboration with Elmar
Wolff in order to avoid manual coding and therefore increasing the scope of the study.
This software program was designed to read pdf. documents and to build databases out
of them. The software also made automatic web searches for bibliographic references.
For example, it makes automatic searches in ISI Web of Knowledge. This software is
under development; consequently it is difficult to evaluate the probability of success.
However, several individual searches were conducted successfully and the automatic
collected data were of comparable quality than the one manually done. Also the
automatic procedure would allow matching information from different online
datasources.
3.2.2. Other sources used to complete CV information
Researchers’ CVs were used as a complementary source of information (see Figure 12).
Information on changes in job positions was obtained through CVs. Data on
publications were obtained through ISI Web of Knowledge. Information on patent
activity was obtained through the European Patent Office. The use of different data
sources helps to avoid the specific limitations of each databases. Questionnaires were
use to gather information on University–Industry collaborations and to complete
information on patent activity. As indicated previously, CVs were used to clean
information coming from ISI and EPO databases.
31
The higher response rate for these disciplines comes also from an earlier request. Previous analysis was
focused on chemistry and physics disciplines. However, higher response rate for chemistry an physics w
tendency in posterior calls.
32
One could also expect the more successful researchers are more willing to send their CVs. However,
we were unable to test this bias towards 'excellence'.
67
Figure 21: Collection methodology
3.2.3. Coding methodology
In order to solve the main problems coming from working with CVs – non-
standardised, truncated and missing information – a first browse of the researches' CVs
was done looking for a common pattern of job mobility data. As a result, researchers
with the basic information on mobility – institution, year, position – were selected. For
the researchers with some data missing, we made an attempt to fill in the missing
information through a web search. A total of 8.7% of the cases were rejected due to lack
of information.
A common problem with the first positions of the researchers is that the institutional
ascription is not very clear. Research assistant positions are sometimes held with a
postdoctoral fellowship, but not all postdoctoral fellowships imply this contractual
relationship. As the information on the CVs was not very clear about this point, the
‘research fellow’ position was considered as the minimum position for the changes in
job position variables. Another variable with the postdoctoral position was created to
keep this information – distinguishing between its national and international character.
Short stays and secondments were not considered due to a lack of consistency in the
information. Some CVs are very detailed, having information on short stays, prizes and
full dates, whereas others only specify educational background and main positions held.
It cannot be assumed that the researchers with limited information on their CVs did not
have any short stays in other university or any governmental position. In addition, date
formats were not consistent (e.g. months are not always reported), making the
information on short stays less reliable.
A careful conceptualisation of researchers' mobility was done in order to reduce the
number of variables and increase the quality of the information. This report only
considered changes in job positions to an institution different from the one where the
PhD was completed. Changes in job position within the same institution were not
68
considered (e.g. a job change in the same university to a higher position). Positions had
to have been for one year or longer. These measures were taken considering that
researchers need time and institutional diversity to acquire new knowledge and
developing valuable networks.
Three conceptualisations of researchers' mobility were considered (see Table 1):
• 'Job mobility' or 'real labour mobility' (JOBMOB). This conceptualisation
considers changes in job positions to an institution different from the one where
the PhD was completed. Positions have to have been for one year or longer.
Each researcher was classified according to the changes in job positions that had
occurred in his or her career. This categorisation – no transitions, transitions to
academia, transitions to industry and transitions to a research centre – was made
according to the maximum heterogeneity of the changes in job position. For
example, if the researcher changed jobs between academia and industry it was
categorised as ‘transitions to industry.’33 However, information on the different
changes in job positions – time, sector and institution – was considered too.
Second and third transitions have two additional variables that consider quality
of changes in job positions (MSTATUSPOS, MSTATUSUNI). These variables
indicate if the researcher acquired a position with higher academic title and/or
got a position in a university with a higher RAE (Research Assessment Exercise)
score through the transition. Detailed information for each change in job position
covers a maximum of three changes in job positions per researcher. This
decision was taken by taking into consideration the fact that almost all
researchers changed position a maximum of three times.
• 'Postdoctoral mobility' (POSTMOB). This variable indicates whether or not the
researcher has held a postdoctoral position. The postdoctoral position has to
have been for one year or longer, and in a different institution to the one in
which the researcher conducted his or her PhD studies. Otherwise, we consider
that the researcher did not use the postdoctoral position as a source of mobility.
Whether the postdoctoral stays in the country of his or her PhD study is also
taken into account.
• 'Broad mobility' or 'collaborations' (JOINTRES, CONTRACTRES,
CONSULRES)34. Some authors35 claim that researchers' mobility should have a
broader conceptualisation than job mobility. This broader conceptualisation of
mobility includes collaborations. Here this broader conceptualisation of mobility
includes university–industry relationships through joint research projects,
contract research and consultancy work. The data of these variables refer to the
U–I collaborations which occurred between 2002 and 2003.
Table 13 presents the structure of the database regarding researchers' mobility variables.
33
This variable includes changes in job position from academia to industry or vice versa.
34
Variables extracted from D'Este et al. 2005 report.
35
Zucker, L.G., Darby, M.R. and Torero, M. 2002. Labor mobility from academe to commerce. Journal
of Labor Economics, 20, 629-660.
69
Table 13: Researchers' mobility variables
Scientists’ real labour mobility- changes in job positions
JOBMOB Variable that indicates if the researcher has changed job positions between different
institutions after her or his PhD. completion.
0- No transitions
1- Transitions to academia
2- Transitions to firm
3- Transitions to research centres
M1TYPE Variable that indicates the sector into which the first transition was made
0- Academia
1- Firm
2- Research Centre
M1YEAR Variable that indicates the year in which the first transition occurred
M1CENTRE Variable that indicates the centre into which the first transition was made
M1POSITION Academic title acquired through the first transition:
1- Professor
2- Reader
3- Senior lecturer
4- Lecturer
5- Senior researcher
6- Researcher fellow
M1TIME Variable that indicates the number of years that the researcher has worked in industry or
Research Centre through the first transition
xxx Xxx
M3STATUSPOS Variable that indicates if the researcher has changed job positions into a higher academic
title through the third transition:
0- Non better position
1- Better position
M3STATUSUNI Variable that indicates if the researcher has changed job positions into a university with a
higher RAE (Research Assessment Exercise) level through the third transition:
0- Non higher institution
1- Higher institution
Postdoctoral mobility
POSTDMOB Variable that indicates if the researcher has held a postdoctoral fellowship.
0- Non postdoctoral fellowship
1- National postdoctoral fellowship
2- International postdoctoral fellowship
Broad mobility- collaborations
JOINTRES Variable that indicates the frequency that the researcher has been engaged in joint research
agreements (between 2002 and 2003):
0- 0 times
1- 1-2 times
2- 3-5 times
3- 6-9 times
4- ≥ 10 times
CONTRACTRES Variable that indicates the frequency that the researcher has been engaged in contract
research agreements (between 2002 and 2003):
0- 0 times
1- 1-2 times
2- 3-5 times
3- 6-9 times
4- ≥ 10 times
CONSULRES Variable that indicates the frequency that the researcher has been engaged in consultancy
agreements (between 2002 and 2003):
0- 0 times
1- 1-2 times
2- 3-5 times
3- 6-9 times
4- ≥ 10 times
70
3.2.4. Characteristics of the sample
The CVs were collected from a total of 807 researchers, 649 of whom had a correct e-
mail address attached to their name. A total of 173 researchers responded by the
deadline and successive reminders. This represents a response rate of 26.66%36. (See
Table 14)
Table 14: Distribution of the researchers by discipline and response rate of e-mails
that reached a valid address
Population Populatio Positive reply Response
n mailed rate
Chemistry 271 212 64 30.19
Physics 195 159 56 35.22
Computer Sciences 162 133 29 21.80
Mechanical, Aero. & Manuf. Eng 179 145 24 16.55
Total 807 649 173 26.66
3.2.5. Analysis
The research questions are listed below, and the type of analysis follows. This section
finishes with a summary of the results and an account of the main methodological
problems encountered.
Research questions
RQ. 1: What are the dynamics of the changes in job positions?
RQ. 2: What is the relationship between the different conceptualisations of mobility?
RQ. 3: What are the characteristics of the researchers involved in job changes in UK
universities?
RQ. 4: Do diverse and increased changes in job positions of university researchers
affect their academic performance?
4.1 When postdoctoral positions are used to support scientists’ mobility, what
effects do they have on productivity (measured by publications and citations)
and career development?
4.2. Does intersectoral job mobility affect academic performance and career
development?
Type of analysis
In order to test the research questions several analyses were performed. Descriptive
statistics, analysis of variances and correlation analyses were use to describe and test the
36
This sample includes start scientists. Some of them are included in the EPRSC sample. That means that
information on collaborations is not available for the whole sample.
71
relationship between different variables. In addition, a selection of mobile and control
group was performed in order to partially control for the main variables that influences
academic productivity – academic experience, discipline, position, RAE score and
department size. The study was exploratory and descriptive in terms of the effect of
postdoctoral mobility and job mobility on academic performance and career
development.
Differences between expected and observed frequencies were used to analyse
postdoctoral mobility and job mobility across research fields and time, and to look at
the relationship between different types of mobility – postdoctoral mobility, and
changes in job positions.
The relationship between researchers' mobility and academic performance relied on
analysis of the correlation between the variables for postdoctoral mobility, job mobility
and precocity on researchers’ performance (RQ. 4.1 and 4.2). Publications, citations and
patents were used as proxies for researchers’ performance.
In the analysis of the correlation between postdoctoral mobility and productivity, three
periods in a researcher’s career are considered (RQ. 4.1). The first analysis was of
researchers’ average publications for a 5-year period starting 2 years after completing
the PhD. Secondly, all career productivity records of a researcher were considered.
Thirdly, the average of publications between 2000 and 2005 is considered. Guarded
against the showing of result by the possible temporal effect of postdoctoral mobility on
academic productivity.
In the analysis of the secoral job mobility (RQ. 4.2), we compared the averages of
cumulative publications of the mobile group – academic and intersectoral – and control
group for a period of time between 2001 and 2005. A study of the yearly publications
between 1985 and 2005 was carried out.
Summary of the results
The analysis of the dynamic of job changes showed an increase in the academic job
transition. The analysis of the relationship between different conceptualisations of
mobility showed evidence of the relationship among them. Researchers who did not
change job position used the postdoctoral fellowship as a source of mobility. Evidence
of positive correlation among previous jobs in industry and a higher U–I collaboration
(broad mobility) was found.
The analysis of the different individual and institutional characteristics showed several
relevant aspects.37 Women were not less mobile than men, but they have a more
traditional pattern of transitions (transitions to academia and research centres). As
predicted, there is a positive correlation between academic experience and age and job
mobility. Being older increases the probability of having more transitions. Transitions
to industry seemed to be a more ‘qualitative’ change as they were not so dependent on
age and academic experience. The analysis of this position showed that professors tend
to move in a ‘traditional’ way.
37
See more details of the results in Zubieta (2008)
72
The analysis of postdoctoral and job mobility showed major differences in the mobility
patterns of basic and transfer scientists. Basic scientists tend to move via a postdoctoral
appointment, whereas transfer scientists tend to change job positions. The sectoral job
mobility analysis also found important differences in the patterns of job mobility across
research fields. Basic scientists tend to change job positions within academia and more
than transfer scientists do. This indicates that it is not possible to apply a general
research policy for encouraging scientists’ mobility.
The analysis of the relationship between researchers' mobility and academic
performance indicates a positive correlation between mobility – postdoctoral mobility
and job mobility – and academic performance. It was found that international
postdoctoral mobility38 is positively correlated with publications for non job-mobile
basic scientists in a 5-year period starting 2 years after the completion of the PhD, and
with the total number of citations.
The analysis of sectoral job mobility indicated a positive correlation between job
mobility and total and yearly average of publication, by comparing productivity of a 'job
mobile' and 'control group' (Graphs 1 to 4, in Figure 13).39 It was found that sectoral
mobility correlates differently across fields; academic changes in job positions correlate
positively with publications for basic scientists, while intersectoral changes in job
positions seem to positively correlate with publications.
In terms of career development, the lack of significance of the results deter this study
from drawing any conclusion. The analysis of postdoctoral positions indicated that
international postdoctoral mobility could be a non-early advantage with positive effects
on a scientist’s productivity and career development. The sectoral analysis of job
mobility indicated that early publications could be a pre-requisite for mobility within
academia. However, it does not seem to be a pre-requisite for moving intersectorally.
38
See more details of the results in Zubieta (2009)
39
See more details in Zubieta 2009 – conference paper.
73
Figure 22: Graphs 1 to 4
Graph 1. Yearly publications per capita ¨Job mobile¨ researchers Graph 2. Yearly publications per capita ¨Job mobile¨
vs. controls: 1981-2005 researchers vs. controls: 1981-2005, by field
Observation range from 42 in 1985 (78 in 1995) and 94 in 2005 Observation range from 42 in 1985 (78 in 1995) and 94 in 2005
Yearly significance: 1998*, 2003*, 2004**, 2005** Yearly significance (basic sciences): 1988*, 2001**, 2003*,2004**, 2005***
*-**-*** Levels of significance at 0.90-0.95-0.99 (Mann-Whitney test for non- Yearly significance (transfer sciences):1999**, 2004**
parametric data) *-**-*** Levels of significance at 0.90-0.95-0.99 (Mann-Whitney test for
non-parametric data)
Graph 3. Yearly publications per capita, transitions to academia Graph 4. Yearly publications per capita, transitions to academia
and transitions to industry vs. controls (basic sciences): 1981- and transitions to industry vs. controls (transfer sciences): 1981-
2005 2005
Observation range from 38 in 1985 (59 in 1995) and 62 in 2005 Observation range from 2 in 1985 (19 in 1995) and 32 in 2005
Yearly significance (academic-control): 1995*, 1998**, 2000*, 2001***, Yearly significance (academic-control): 1997**, 2004**
2003**, 2004**, 2005*** Yearly significance (academic-intersectoral): 1991**, 1992*
Yearly significance (academic-intersectoral): 1996*, 1998*, 1999**, 2005* Yearly significance (intersectoral-control): 1987*, 1990*, 1992**,1999*
*-**-*** Levels of significance at 0.90-0.95-0.99 (Mann-Whitney test for non- *-**-*** Levels of significance at 0.90-0.95-0.99 (Mann-Whitney test for
parametric data) non-parametric data)
Methodological problems
The main methodological problem comes from the sample size. A small sample size
makes it difficult to perform an analysis controlling for all the variables across a long
period of time. The endogeneity problem, cohort and vintage effects are hardly
controlled in an analysis on small sample sizes.
74
4. CV database mapping in other European and associated countries
4.1. FRANCE
Isabelle Recotillet, Observatoire des Science et des Techniques.
The main report that can be made for the French case is that there is neither centralised
nor harmonised data sources for CVs. This situation can be partly explained by the
specific structure of the French research and academic system. A lot has been written on
that point for instance by Musselin (2005) and Laredo and Mustar (1998). We will now
mention some particular points for a better understanding of the French situation.
The main feature of France concerning higher education and research is the institutional
barrier between the two. The former has been set up quasi-independently of the latter.
At the beginning, research activities were carried out mainly by the CNRS (National
Scientific Research Centre) and by some other thematic research institutions (INRA:
National Agronomic Research Institute; INSERM: National Medical and Health
Research Institute; IRD: National Development Research Institute…), whereas
universities were in charge of education and training. Until the middle of the 1990s, the
faculties were discipline-based and “each of them could thus adopt their own pattern of
internal regulation” (Musselin, 2001). The situation has slightly changed and mixed
laboratory units between CNRS and universities have been settled.
To sum up, in France, research activities are carried out by several actors. This has
consequences on careers first and then on the centralisation of information about careers
– of which the CVs is a major component.
Nevertheless, a fruitful attempt was conducted in 2004 by researchers from INRA
(Sabatier, Carrere, Mangematin, 2006)40 based on registers and CVs from INRA which,
to this day, remain an experimental work. The main objective was to better understand
scientists’ careers, focusing on those in biology as a case study. The database was built
using three different sources from INRA. On the 2,200 researchers employed by INRA
in 2003, 583 constitute the sample of the CV database, for which information is
provided by “different datasets [that] have been matched in order to describe their
careers and scientific activities. The first dataset is administrative and it describes the
speed of a career, i.e. the time spent as a researcher before being promoted to professor.
The second dataset comes from INRA’s human resource management office, which
collects every two years the curriculum vitae of its researchers to assess their activity.
Curriculum vitae were used to characterize the activity profiles of researchers, from
year of recruitment to a researcher position at INRA to the time of promotion to
professor, or to the end of 2002 if they were still in a researcher position at that time.
[…] In order to have complete information about researcher’s publications, we matched
our data with a third dataset” (Sabatier, Carrere, Mangematin, 2006). The third source is
a sample of the Science Citation Index (SCI), which informs about the number of
publications per researcher from 1990 to 2002 and on the quality of the journal.
40
Sabatier, M.; Carrere, M. y Mangematin, V. Profiles of Academic Activities and Careers:
Does Gender Matter? An Analysis Based on French Life Scientists’ CVs. Journal of
Technology Transfer, 2006, vol.31, 311-324.
75
Finally, the sample allows the following of the career of researchers from their PhD till
2002. The administrative part of the dataset brings information about: year of
recruitment, year of promotion, year of defence of the “habilitation à diriger des
recherches”, managerial responsibilities and professional affiliations. The CVs provide
gender, age, university of graduation, geographical mobility before the recruitment
(post doc for instance) and after (visiting mobility for example), as well as relevant
indicators on research activities. As stated previously, the third source provides in
addition the knowledge about publications.
This is the main source available in France. It is the most relevant to our research.
We can easily assume that the other research institutes collect CVs and different
information about their researchers in the context of assessment procedures. However,
we do not have access to this information.
A potential source for the analysis of the beginning of careers of young researchers
could be the CV data source collected by the Bernard Gregory Association (ABG,
http://www.abg.asso.fr/). About 1000 CVs are stored by the Association, those of young
PhD graduates looking for their first job into companies. The main objective of ABG is
to promote the placement of PhD graduates and young researchers into the non-
academic sector. It would offer the possibility to explore the career paths of researchers
from PhD to a job in the private sector. However, it offers a low level of
standardisation, which could be a serious limitation for analysis, even if the CVs are
built on the same pattern: degrees, thesis, skills, work experience, languages and
geographical mobility. For each topic, the candidate is free to fill in what s/he wants.
Anyone can access the CVs online using the website.
4.2. ISRAEL
Daphne Getz, Bella Zalmanovich & Tsipy Buchnik, Samuel Neaman Institute, Haifa.
4.2.1. Introduction
Higher education in Israel is under the direct jurisdiction of the Council for Higher
Education, which is responsible for accrediting and authorizing institutions of higher
education to award degrees. The institutions of higher education are autonomous in the
conduct of their academic and administrative affairs within the framework of their
budgets. In the academic year 2008/09 there were 67 institutions of higher education
including: 8 universities, 30 academic institutions which are not universities (publicly
and non-publicly funded), 27 colleges for training teaching staff, and academic tracks
under the aegis of academic universities in 2 regional colleges.
The following seven universities engage in both teaching and research: the Hebrew
University of Jerusalem, Technion–Israel Institute of Technology, Tel Aviv University,
Bar-Ilan University, the University of Haifa, Ben-Gurion University of the Negev, and
the Weizmann Institute of Science (a research institute that offers graduate programs).
In addition to the universities, a large variety of institutions are accredited as institutions
of higher education and offer academic programs in a broad spectrum of fields, such as
the fine arts, business, law, music, technology, and teacher training.
76
The advancement of human knowledge through research is a central task of the research
universities in Israel. Therefore the universities are the natural environment for students
who are interested in research and the training of PhDs in Israel is done in the seven
research universities. Young researchers completing their PhDs are encouraged to
complete their post docs studies abroad and senior researchers are encouraged to stay
abroad for sabbatical. Israel also participates in the Euraxess – the European Union
program that assists researchers moving between countries.
4.2.2. National Level
The following institutes systematically collect Researchers’ CVs in Israel:
1. Israel Science Foundation (ISF) collects CVs with the application for different type
of grants. The applicants send their proposals via the ISF’s online electronic submission
system. The CV screen includes four parts: Academic Background, Previous
Employment, Grants & Awards and List of Publications. The researchers CVs are
stored only in the ISF database and are used only for the ISF grants applicants’ selection
purposes. With every new application, the researcher updates his CV.
The Israel academy of sciences & humanities is planning to collect researchers CVs
within the project of establishing a register of young Israeli scholars and scientists
working/studying outside Israel.
2. Higher Education Institutes: The 8 Universities & the rest of the Academic
Institutes collect researchers’ & academic staff CVs.
The information detailed below will focus on the researchers’ CVs at the Israeli
research universities. Since the universities are the centres of research activities
performed in Israel and therefore are most relevant to the Euro CV project.
4.2.3. Institutional Level CVs collection in the Research Universities
Currently, there is no formal policy regarding the collection and storage of researchers’
CVs. The researchers’ CVs can be uploaded to their organizational unit (faculty,
department, etc.) homepage or to their personal homepage. Uploading the researcher
CV to organizational unit homepage is recommended but not mandatory.
Over the years researchers have begun to understand the importance of the Internet.
Today more researchers are uploading their CVs to the Internet. Young researchers tend
more to upload their CVs or to build their personal homepage. The tendency to upload
CV’s depends also on the organizational unit policy/traditions. The storage formats of
CVs are varied: pdf, html and word files.
CV Information can also be retrieved from the universities Research & Development
Authorities. The authority for research and development is the main administrative
department responsible for encouraging, promoting and organizing research activities at
the university. Each university research & development authority has its own register
77
containing a list of researchers and their research grants (as presented in Table 15). The
CVs information on the registers is partially updated: for example, the publication list
covers selected number of publications mostly from the 1980’s and forth.
Table 15:
Research *Number of Fields included Missing fields
Authority researchers
Name included
Hebrew http://www.huji.ac.il/d 2,442 (by personal details; academic Employment history;
University of ataj/controller/ihoker family name) degree; academic positions; full publication list;
Jerusalem – administrative positions; fellowship, honours
The Authority external academic and awards;
for Research positions; research grants;
and research interests; research conferences
Development projects;
selected recent; publications
Technion http://www.trdf.co.il/E 1,078 (by faculty; personal details:
Research & ng/redir.asp?PageId=R faculty name) research interests; research education;
Development nD/researcher projects full publication list;
Foundation academic positions;
external academic
positions;
research grants;
fellowship, honours
and awards;
conferences
Haifa http://ra.haifa.ac.il/ra1. 1,610 (by personal details; academic positions;
University – asp?page1=Research_ family name) research Interests; research external academic
Research Staff.asp disciplines; positions;
Authority higher education; full publication list;
recent publication; courses; research grants;
keywords fellowship, honours
and awards;
conferences
Bar Ilan http://research.biu.ac.il 592 (by personal details; education;
University - /?p=1 faculty name) research interests; academic positions;
Research publications; external academic
Authority key words positions;
full publication list;
research grants;
fellowship, honours
and awards;
conferences
Ben – Gurion http://profiler.bgu.ac.il 721 (by contacts education;
University - /research/ faculty name) research interests; research academic positions;
Research projects; external academic
Authority recent publication; positions;
keywords full publication list;
research grants;
fellowship, honours
and awards;
conferences
*The number of researchers includes emeritus professors.
78
4.2.4. Researchers CVs Format
There are two main formats as detailed below:
Standardized CV format
There is a recommended standardized CV format for faculty members. This format is
usually used for appointment, promotion or selection purposes. The standardized format
slightly differs among the different universities. The fully standardized researcher’s CV
format (taken from several universities) includes the following fields:
- Personal/Contact Details (name, place & date of birth, marital status, citizenship,
address, telephone, fax, e-mail, personal home page);
- Education (Academic Degrees, Clinical Training);
- Academic Appointments;
- Administrative Appointments;
- Visiting Appointments;
- Professional Experience/Employment History;
- Other Appointments;
- Teaching Experience;
- Public Professional Activities;
- Research Interests;
- Membership in Professional Societies;
- Fellowship, Honours and Awards;
- Research Grants;
- Publications (Theses, PhD Dissertation, refereed papers in professional journals,
original papers, review papers, book chapters, books and special Journal Issues
edited, patents, research reports, other publication);
- Conferences;
- Special Professional activities;
- Graduate Students;
Electronic Template
The template for electronic CV’s is used for presenting the researchers’ CVs on the
internet. This template differs slightly among the universities and among the different
faculties, but basically it includes the following fields:
- Personal/contact details;
- Research Interests/areas;
- Education (usually from the PhD level);
- Recent Publications (from the last ten years);
- courses
The standardized CV format is more comprehensive, detailed and includes complete
information on the researcher’s careers and scientific outputs (for example: full list of
publications, grants, fellowships, awards, memberships, activities, etc.). The electronic
format is more concise and includes the researcher’s main or recent publication and
activities.
79
4.2.5. SNI’S Selected Registers for the Prime Euro-CV Project
20 researchers' CVs were collected from the Racah Institute of Physics at the Hebrew
University of Jerusalem and from the Physics Department at the Technion - Israel
Institute of Technology. The researchers were selected randomly from the following
databases: The Hebrew University of Jerusalem - The Authority for Research &
Development Register41 and from the Technion Research and Development Foundation
Register.42
For the coding process we used three sources of Information (as presented in Table 16):
Table 16: Sources for coding CVs
Source Fields included Comments
The Authority for Personal Details; Academic Degree; Academic *Free access
Research & Positions; Administrative Positions; External *Database format
Development Registers Academic Positions; Research Interests; Research
Projects; Selected Recent Publications
Number of cvs
CVs - electronic Personal/contact details; Research A concise format used
template from the Interests/Areas; Education; Recent Publications for displaying faculty
faculty homepage members CVs on the
Internet.
*Free access
*HTML format
Full CVs uploaded by Personal/Contact Details; Education (Academic Standardized CV format
the researchers Degrees, Clinical Training); Academic designed for faculty
Appointments; Administrative Appointments; members or applicants
Visiting Appointments; Professional for faculty positions.
Experience/Employment History; Other This format is usually
Appointments; Teaching Experience; Public used for appointment,
Professional Activities; Research Interests; promotion, or selection
Membership in Professional Societies; purposes.
Fellowship, Honors and Awards; Research *Free access
Grants; Publications (Theses, Ph.D dissertation, *DOC or PDF format
referred papers in professional journals, original
papers, review papers, book chapters, books and
special journal issues edited, patents, research
reports, other publications); Conferences; Special
Professional Activities; Graduate Students;
Table 17 presents the information which was coded.
Table 17: Coded information
Field Remarks
Personal Details
Birth Date
Family status
41
http://www.huji.ac.il/dataj/controller/ihoker - The Hebrew University Research & Development
authority Register covers Hebrew University researchers from all disciplines who applied to the research
authority for grants and/or funds, includes approximately 2,000 items.
42
http://www.trdf.co.il/Eng/redir.asp?PageId=RnD/researcher - The Technion Research and
Development Foundation Register covers Technion researchers from all disciplines who applied to the
research authority for grants and/or funds, includes approximately 1,000 items.
80
Contact details (address, phone, fax, e-mail, url)
Academic Details
Institution
Discipline
Research Interests
B.A Degree
Organization
Date
Country
M.A Degree
Organization
Date
Country
Phd Degree
Organization
Date
Country
Visiting Positions
Status
Organization
Starting date Only year
Starting date Only year
Job Positions
Organization
Position
Starting Date Only year
Finishing Date Only year
Articles
Title
Author(s)
Journal
Year
Volume
Pages
Books
Title
Author(s)/
Year
Pages
Book Chapters
Title
Author(s)/Editor(s)
Year
Pages
Patent
Title
Author
Date
Institution/country
Research Grants
Title
Amount
Country
Funding Org
Year
Conferences
Name
Year
81
Country
Fellowship/Awards/Prizes/Honors
Name
Year
Country/Institution
Supervised PHDs students
Name of student
Theses title
Year
4.2.6. CV Coding Issues
CV updating – No date is listed on the CV electronic template. Therefore the CV’s
updating frequency is unclear.
CV scope – The information appearing in the CV is not necessary comprehensive and
unified. For example, the following information is partly missing:
• Purpose of the researchers mobility – is the mobility intended for post doc,
sabbatical or appointments purposes.
• Specification of the mobility period – particularly for mobility periods that last less
than a year.
• Full list of publications – usually in the CV electronic template, the selected list of
publications includes publications from the last two decades
• ISBN of books, ISSN and Impact factor of journals fields usually doesn’t appear in
the researchers’ CVs.
To complete the CV’s missing details, we used the following databases:
• Bibliometrics databases – ISI web of knowledge, Scopus
• Patent databases – U.S & European patent offices
• Personal appeal to the researcher should be considered in order to verify the update
and the completeness of the CV.
Additional Issues:
Israel is country which encourages and absorbs Jewish immigrants from all over the
world. During the 90's, approximately 800,000 Russian Jews immigrated to Israel,
among them a high percentage of researchers, scholars, doctors etc. Special programs
were formed in order to assist them to in finding employment in institutions of higher
education, research institutes, technological incubators, and in the industrial and
business sectors. Therefore, when coding researchers CV’s born outside Israel, the
following questions arise:
• Coding of academic positions held prior to the immigration – are these positions
considered part of the mobility.
• Discrepancy between academic positions and academic degrees in Israel and
abroad.
4.2.7. Policy recommendations
In conclusion, the current coding process is inefficient and takes long. Therefore, we
suggest the following recommendations:
82
• Updating the information on the research & development authorities registers on
a regular basis. Currently, the information stored there is not updated and
includes retired and descend researchers.
• Creating a unified and comprehensive CV format as possible for researchers.
Currently, the CV format is slightly different between the universities.
• At the national level - creating a unified and standardized database that will
include researchers’ CVs from all the research & development authorities at the
Israeli universities. In the past, the Forum of University Research Authority
Directors (FURAD) tried to establish such a database in order to ease the
process of allocating research partners in Israel. However, this project was not
completed. This project faces two main obstacles: achieving the appropriate
funding and creating collaboration between the universities.
References
Israel Council for Higher Education
http://www.che.org.il/
Israel Science foundation
http://www.isf.org.il
Israel Science foundation online User Guide
Report of the Committee for the Examination of the system of HIGHER Education in
Israel
http://www.che.org.il/download/files/%D7%93%D7%95%D7%97_%D7%91%D7%99
%D7%99%D7%92%D7%94_%D7%91%D7%90%D7%A0%D7%92%D7%9C%D7%9
9%D7%AA.pdf
4.3. SWITZERLAND
Carole Probst and Benedetto Lepori, Università della Svizzera italiana, Lugano.
4.3.1. Introduction
Switzerland is a federal State in which decisions tend to be taken on the lowest political
level possible. Regarding higher education, authority is spread between the
Confederation and the Cantons. While the two Federal Institutes of Technology (FIT)
are under federal authority (but dispose of large autonomy), the universities are under
cantonal authority; therefore, no central body with the power to regulate higher
education and higher education institutions in the same way all over the country exists.
Coordination occurs by consensus, most often through two bodies: the Swiss University
Conference and the Rector’s Conference of the Swiss Universities.
In this context, it is interesting to understand to what extent common databases
containing information on researchers’ careers and mobility, thus CV-like information,
exist, and whether such databases can be used for research purposes. This is the aim of
this study. The text at hand first addresses the general situation regarding the storage of
83
researchers’ CV information in databases in Switzerland. Then, the situation at the
individual universities is presented. Finally, results of the analysis of the 20 CVs as well
as observations from a previous study are presented. The text is concluded with some
remarks on possible and desirable future developments.
The information for this report was collected in three steps. First, people responsible for
research services at Swiss universities were contacted and asked whether at their
university a central database containing CV-like information exists, and whether they
know about any such database outside their own university. Then, the websites of all
universities were closely looked at in order to identify databases and patterns at the
level of the whole university and of the single departments and institutes. As a third
step, a random sample of 20 CVs from two universities was retrieved and analysed
according to their structure. Data and information was retrieved in spring 2008.
4.3.2. General situation
National level
The Swiss National Science Foundation (SNSF) collects CVs with the application for
different types of grants. These CVs are stored in paper form in the dossiers. Electronic
submission has been introduced only recently, and is limited to the submission of files
(PDFs) containing CVs. With every new application, a researcher re-submits his CV.
The CVs of the researchers that have applied for research grants at the SNSF are stored
in the respective dossiers in paper format. Some exploitation of these CVs has been
done recently in the context of two research projects:
A study on “Women in research” commissioned by the SNSF is currently ongoing
(SNSF 2006). This study starts with an analysis of information on individuals contained
in the Swiss University Information System, uses a survey on graduates, with particular
interest in their behaviour regarding applications to the SNSF and analyses the
application system of the SNSF. Then, CVs and publication lists in this system are used
for analysing further gender-specific differences in individual scientific profiles of
selected groups of successful and unsuccessful applicants. Finally, interviews complete
the study. In this project, the analysis of CVs and publication lists is used as one
element for analysing gender-related differences in careers.
Another study evaluated the SNSF Professorships programme (Goastellec et al. 2007), a
funding programme aimed at young researchers after a post-doc period of at least two
years. Besides other data sources, this study also relied on information retrieved from
the applications to the programme, thus from the submitted CVs. Information retrieved
from the applications included gender, age, nationality, scientific domain, hosting
university, start, end and amount of funding of the professorship.
The Rector’s Conference of the Swiss Universities CRUS has set up a common
database, www.proff.ch, where all professors employed at Swiss universities are
included. This database, however, contains only contact information, institutional
affiliation, research area and links to the units’ websites.
84
Institutional level
In order to understand whether and how CVs are stored at the single higher education
institutions, we have contacted representatives from research services or other similar
units from all ten cantonal universities plus the two Federal Institutes of Technology
and asked them whether, at their institution, a central database exists, and if yes how it
is used.
The answers show that no university or FIT in Switzerland currently uses a central CV
database. Some smaller databases, for example for publications or projects, are in use at
some places, but heterogeneity is high. In the following, an overview on the situation is
provided.
Most universities’ homepages offer a directory where the visitor can search for people.
In some directories, it is possible to get lists of all collaborators (for example by
searching for a*, b*, etc.), while in others one needs to know at least three letters or the
complete surname of a person, thus it is not possible to use this directory for creating
lists of all collaborators. The groups of people included in these directories differ; in
some cases even students of the university are listed.
Often, these directories, accessible usually from the first page of the websites, include
only contact information and no further information on the person or a link to other sites
providing more information.
Half of the universities – including the two federal institutes of technology ETHZ and
EPFL, the three rather small universities in Neuchâtel, St. Gallen and the Italian
speaking part of Switzerland as well as the university of Lausanne – have a central
directory including more than just contact information. These databases allow people to
introduce information on their background, research areas, publications, mandates, etc.
It seems, however, that it is in the researcher’s responsibility to introduce the
information, and therefore the databases are far from being complete and contain
different types of information in terms of content and structure (lists, text, etc.). In
addition, the categories that can be filled in are rather generic (such as training,
professional experience, etc.) and thus can be and are interpreted in different ways – the
contained information is not homogeneous at all. Information is usually available
directly on the website, but also PDF-files containing CVs and publications are used,
usually in addition to the information contained on the website. In some cases, it is
possibly to automatically generate PDFs from the database (Neuchâtel, St. Gallen).
Standardised databases can be found in some universities regarding publications. But
also in this case, it depends on the researcher to introduce information, and thus it seems
that there are many incomplete lists.
Especially in the universities without central databases with more than contact
information, researchers often use their personal pages or pages of the organisational
unit to present CV-like information. Here, heterogeneity is even higher. While already
the pages of the organisational units differ in many universities – in terms of structure,
content, but also design, it often seems that there is no corporate design at the
universities or it is not respected – several professors link to their personal homepage
outside the university server for more information, containing even pictures of their last
holidays.
85
It seems that also the organisational unit has an influence on whether people publish
information on themselves on the website or not. In the universities without central
directories, there are organisational units that include information on every researcher,
or at least on every professor, and in some cases it seems that internal databases of the
units exist. Rather complete information is usually found at the faculties of law, where
often directories at the faculty level exist. At the other extreme are the sciences – here,
websites give mainly information on research projects and publications, but information
on individual people is rare – if year and area of the doctoral degree or the position at
the university are included, it’s often already a lot.
So overall, there is a tendency to upload CV information and publication lists on the
internet, but this seems to depend a lot on traditions and habits within the organisational
units (faculties, institutes). Overall, however, the heterogeneity of the available
information is large, there is no standardisation at all. At some organisational units, the
information provided looks similar for several researchers, but it seems that this occurs
rather because people copy from each other than because of institutional guidelines.
Some more standardisation can be found at those six universities with a central
directory containing more than contact details. Here, at least the categories are given.
They are, however, rather large, and people are free to decide which information they
want to provide and whether this information is visible to the public or not.
Thus, analysis based on this information always entails restrictions. The group of people
on which information is available is probably biased – who are the professors that do
not publish information on themselves? – and the contained information is
heterogeneous. It is not possible to use the databases directly for analysis, but manual
coding is necessary in order to retrieve at least some information. It seems, however,
that when concentrating on a sample size that allows for manual coding, it is possible to
get at least some basic information on milestones in the researchers’ careers (doctoral
degree, positions hold for a longer period).
4.3.3. Universities without central databases containing CV-information
In the following, the situation at the 12 universities is shortly presented. Information is
retrieved from the Internet as well as through direct contacts with people responsible for
research services at the universities.
University of Basel
There is no common database at the university of Basel. A search function that allows
searching for teachers is available on the website, but at least three characters of the
name have to be introduced.
Recently, a new research database has been created. However, the responsible people
decided not to include structured CV information, as this information is already
contained in other places, as the websites of individual researchers or units, and as this
would require maintenance by the individuals. Thus, information can be found only at
the faculty (in 2 cases: theology and law) or institute level. Every organisational unit
(faculty, department, institute, group…) has its one homepage (they differ in terms of
86
layout, structure, content). In many organisational units, information is available only
on a (small) sample of researchers.
There is no standardisation, it seems that it is up to the researcher to include the
information. The highest degree of standardisation can be found in the faculty of law,
where for all professors a profile (including contact information, a selection of
publications and collaborators), publications and curriculum vitae are included; the
provided information, however, varies. The publication list seems to be standardised,
probably based on a publication database.
University of Bern
There is no common database with CV or publication information at the university of
Bern, and there is no such database planned so far. It is possible to search for people on
the website, but one needs to know the exact name, and the results give only the contact
information and no link to more information on the person.
Generally, the homepages of the different organisational units are in the same system
and thus look rather similar – with some exceptions. Especially at the Faculty of
Medicine, most departments/groups have their own homepages.
CV-like information and some information on publications is contained for most
collaborators in all faculties. The structure of the information often seems to depend on
the organisational unit – often, the structure is rather similar within CVs from the same
institute. In the institute of philosophy, for example, for every researcher, information
on the background (CV-like), on research interests and a downloadable publication list
are contained. In the language departments, CV-like information is often written in text.
At the faculty of sciences (philosophisch-naturwissenschaftliche Fakultät), CV-like
information is less common. Publications (sometimes also in sections containing the
publications of the whole team) and research areas however are often contained.
There seems to be different “CV publication cultures” according to disciplines. Do you
think this would be worth mentioning? I had already included something in the
introductory paragraph on the institutional databases (p. 3: “Rather complete
information is usually found at the faculties of law, where often directories at the
faculty level exist. At the other extreme are the sciences – here, websites give mainly
information on research projects and publications, but information on individual people
is rare – if year and area of the doctoral degree or the position at the university are
included, it’s often already a lot.”) As my analysis however is rather small, I would not
include more than this.
University of Fribourg
There is a directory of the university, but one has to know at least the first three letters
of the name of a person. The results give only the contact information and sometimes a
link to the organisational unit the person belongs to.
There is no common database including CVs, but a research database (Futura -
http://admin.unifr.ch/futura/faces/pages/index.xhtml). In this database, the members of
every research unit are displayed, but there is only contact information and no link to
87
additional information. Research projects and publications are included as well. This
database could probably be used for analysis of publications.
The faculties of law and science offer a list with all their professors, including links to
individual websites. The individual websites, however, differ a lot, and some are located
on domains outside the university. Information varies as in the other universities – from
one phrase to complete CVs and publication lists downloadable as PDF files.
University of Geneva
The central directory (Annuaire) contains only contact information, no further
information or links to personal sites. There is a central publication database, but it
contains only a selection of publications per person and does not seem to be updated
regularly by every researcher.
In some faculties or units – for example law, theology or the sociology department in
social sciences – it seems rather common to upload CVs and publication information on
the internet.
In other faculties, it depends heavily on the organisational unit. The directory of people
from the science faculty links to a national database containing only contact
information. Information on research areas and publications, also of whole teams, is
more common than CV-like information.
University of Lucerne
There is a central directory containing contact information, but no further information
or links.
In the websites of the faculties and institutes, information on nearly every professor is
available, usually in a rather complete way, containing CV information, information on
research and publications.
At this rather young university – established when Internet already existed - the
websites of every organisational unit in all faculties are included in the same general
website, with the same general design. It seems that there is some policy or it is just
common to upload CV, research and publication information on the websites.
University of Zurich
Also in Zurich, the central directory contains only contact information and no links or
additional information.
Information can be found on the websites of the individual organisational units. These
websites are not included in the same design, and they vary a lot. Also the completeness
and type of information provided regarding researchers’ CVs varies.
In the law faculty, for example, information on most professors is provided, on websites
or as downloadable PDF-files. In the sciences, information on research areas and
publications is more frequent than information on the curriculum.
In the larger faculties, information is contained in the homepages of the sub-units, but in
a very heterogeneous way.
88
4.3.4. Universities with a central database containing CV-information
EPFL - http://people.epfl.ch/
At EPFL, there is a central database that allows researchers (but also other members of
EPFL such as students, co-workers) to introduce information including CV,
professional history, publications, work, skills, etc. This can be used by individuals, but
also by groups. Everyone can choose which part of the information is visible. This
database was originally implemented in order to produce yearbooks of teaching staff.
The directory can be searched online.
This tool, however, is not used by all staff members, and it is used in very different
ways. When looking for example at the site of the president of EPFL, only a short text
called „Biography“ appears. Others include information on publications, on current or
recent research projects, etc. Most sites, however, include only information on the
person’s affiliation or only some information in a few categories.
ETHZ - http://www.ethz.ch/people/whoiswho
At ETHZ, a central directory including all 521 professors exists: The Who’s who. It can
be searched on the Internet.
For every professor, there is a written text containing biographical information. Its
length varies, and there is no standardisation in it. A list of taught courses is contained,
and links to databases on events and publications are provided.
In many cases, a link to the personal homepage is provided (or at least to the website of
the organisational unit). There, in some cases more information is available, for
example publication lists or CVs downloadable as PDFs.
University of Lausanne - http://www.unil.ch/unisciences
There is a central directory containing information on contact, curriculum, research,
teaching and publication – according to what is filled in. Level of detail varies, not all
researches provide in complete information. It seems that a publication database
automatically generates the contact information, publication lists, and teaching section.
Regarding curriculum and research areas, many researchers do not provide information.
The section “Curriculum” contains information on training, other activities, professional
experiences, usually as lists, thus there seems to be some standardisation.
This database could probably be used for some analysis, but it is limited due to the fact
that in many cases no information is contained. Also the publication database seems
interesting for analysis.
University of Neuchâtel - http://hydra.unine.ch/cvprof/
There is a database containing information on professors and teachers. Contained
categories are more detailed than in other databases, but still remain rather broad (e.g.
professional life, university degree, main publications). The level of completeness
89
varies – not everybody has filled in all categories. Information from this database is
included in the websites of some organisational units, but not in all.
From this database, it is possible to automatically generate PDFs of the contained CVs.
To use this database for analysis, manual coding would be necessary even if it was
possible to access the information in database format: the categories are rather broad
and thus the provided information varies.
University of St. Gallen - http://www.alexandria.unisg.ch
There is a central directory (Alexandria) that contains information on all collaborators.
On the websites of the individual units, people are usually listed, but often there is no
direct link to their entry in Alexandria. In some institutes, information from the
Alexandria database is directly integrated in the websites presenting the researchers.
Not for all collaborators, however, every available category contains information, and
the way in which details are filled in varies as well. There is no standardisation inside
the rather broad categories.
The database includes also all publications and projects, here the level of
standardisation seems high. This database would probably allow for interesting analysis
of publication patterns, also because it seems that it provides rather complete
information.
Università della Svizzera italiana - http://www.unisi.ch/people_directory.htm
There is a central database in which all collaborators are included, accessible from the
“search person” function on the main website.
This database contains links to projects and courses – it seems that they are generated
automatically – as well as sections on CV and publications.
CV information and publications have to be inserted by the people themselves, and thus
the database is not complete and varies in the levels of detail. There are no sub-
categories, all CV-information is contained in the same field. Usually, CVs are provided
as written text. Some researchers also include their CVs and/or publication lists as PDF
files. The publication list has a rather high degree of standardisation.
University available fields Completeness of CV standardisation of CV
containing CV information information
information
EPFL A broad range of fields, From zero to rather complete Several fields, no
including academic information, depending on standardisation within
background, researcher
professional
qualifications,
biography, research
topics, publications
ETHZ Curriculum Vitae (1 Provides basic information for No standardisation, only 1
field), link to every researcher (about 200 field
Publications Database words per researcher)
90
Lausanne curriculum (other From zero to rather complete 3 fields in curriculum
activities, work information, depending on category, no standardisation
experience, education), researcher within, but often list.
research, publication
Neuchâtel A broad range of fields, Some information seems to be Several fields, no
including professional contained for most researchers standardisation within
life, university degree,
main publications,
affiliations, etc.
St. Gallen Earlier positions (most Often rather scarce, besides Several fields, no
often used field), automatically generated standardisation within
memberships, other positions concerning actual
information position
Università Biography (1 field), From zero to rather complete No standardisation, only 1
della publications information, depending on field
Svizzera researcher
italiana
4.3.5. Available information and possible analysis
In order to understand what type of information is generally available in databases or
other forms of presentation on the institutes’ websites, a small sample consisting of 20
CVs was analysed. The sample contains 10 CVs from the university of Lucerne – a
university without a central database, but with rather complete information for every
professor – and 10 CVs from the university of Neuchâtel, a university with a rather
detailed central database that covers at least some basic information for most professors.
The CVs were randomly chosen among all departments and organisational units.
Table 18 shows which type of information is available, both overall and at the two
different places. Bold characters indicate categories on which much information is
available, while high numbers of missing information are indicated in grey.
Table 18: Curricular information available in the databases of the University of
Lucerne and the University of Neuchâtel
overall Lucerne Neuchâtel
yes no yes yes comments
Age (year of birth) 13 7 7 6
Gender 20 0 10 10
Marital Status 7 13 2 5
Discipline 20 0 10 10
Citizenship 2 18 2 0
Residence 5 15 1 4
Children 5 15 2 3
research interests 17 3 7 10
PhD degree
Organization 15 5 8 7
date 13 7 9 4 Date: only year
country 15 5 8 7
Visiting Scholarships
organization 9 11 7 2
starting date 9 11 7 2 Dates: only year
91
finishing date 8 12 6 2
country 9 11 7 2
job positions
organization 19 1 9 10
position 19 1 9 10
starting date 14 6 9 5 Dates: only year
finishing date 13 7 8 5
country 19 1 9 10
research grants
title 5 15 5 0
amount 0 20 0 0
coordinator 0 20 0 0
country 5 15 5 0
funding org 5 15 5 0
supervised phd
organization 0 20 0 0
name of student 0 20 0 0
discipline 0 20 0 0
country 0 20 0 0
grade 0 20 0 0
patents
number 0 20 0 0
date 0 20 0 0
country 0 20 0 0
NE: only selection of
publications (5 per
Books researcher)
title 15 5 9 6
author(s) 15 5 9 6
nr. pages 3 17 2 1
ISBN 0 20 0 0
year 14 6 9 5
place 12 7 8 4
book chapters
title2 11 9 9 2
author(s) 9 11 7 2
nr. pages 6 14 6 0
ISBN 0 20 0 0
year 11 9 9 2
place 10 10 9 1
Articles
title 15 5 9 6
journal 16 4 9 7
author(s) 13 7 7 6
volume 13 7 8 5
year 17 3 9 8
month 0 20 0 0
nr. pages 13 7 8 5
ISSN 0 20 0 0
impact factor 0 20 0 0
92
This table shows that there is absolutely no information available on supervised PhDs
and on patents. Regarding publication, ISBN and ISSN, the month and impact factor for
journal publications are always missing. When information on research grants is
contained (five cases from Lucerne), there is no information on the coordinator and on
the amount.
Information on gender is not explicitly contained, but can be deducted in all cases from
the names or from titles indicating the gender (“Professorin” vs. “Professor”).
Information on age, marital status and children seems not to be considered as a
constitutive part of a CV, and explicit information on citizenship is contained only in
two cases. Also residence is often missing.
As there is no standardised CV format, missing information can be interpreted in two
ways. Or it was not inserted in the CV (because it was forgotten, deliberately omitted,
considered as not important), or the event never took place – for example, a researcher
never had a research grant, never was a visiting fellow at another university, or never
published a book. Also where information is included, it is not possible to know
whether this information is complete – for example whether all positions and all visiting
scholarships are contained or just a selection is presented.
With data from this kind of CVs, it would be possible to trace a researcher’s mobility
regarding permanent positions, starting with the PhD. With data from Lucerne, where
publication lists seem to be rather complete, one could also trace a researcher’s
publication activity, in terms of quantity, but also language and place of publication,
during his career. One could also analyse the words used in titles of publications, and
thus analyse the development of the research topic. Also mobility between academic
and non-academic employment could be analysed. This kind of mobility seems to be
particularly common in fields such as law, where professors often work as lawyers as
well, or theology, with professors also having appointments as priests.
Given the high coverage of these two databases and the restricted dimensions of the two
universities, it would be possible to analyse the general mobility in terms of stable
positions of the whole population of professors, which would allow giving an overview
on the overall background of the professors.
Detailed analysis of mobility, however, is difficult to do with Swiss CVs: information
on the dates of stays abroad or stable position is usually restricted to the year, and thus it
is not possible to deduct the exact duration of these stays. It also seems that shorter
stays are not always included in the CVs.
In a previous study conducted in 2006 (see Lepori & Probst 2008), we used CV and
publication information retrieved from the Internet as data sources. This study has
shown that, even though a high degree of heterogeneity prevails, it is possible to use
this type of information for specific purposes. The aim of that study was to map the
field of communication sciences in Switzerland. CVs and publication lists were used as
major sources of information. In that case, it was possible to retrieve at least some
information for all but one of the 67 professors in the field. Overall, 28 more or less
complete CVs, 47 short descriptions containing CV-like information and 53 publication
lists could be retrieved. We constructed the coding procedure for CVs and publication
lists according to the available information and according to our purposes (to construct
93
a general map of the field). Regarding the CVs, we mainly included information on the
current position as well as on the first degree and the doctoral degree (place, year, field).
We thus used CV information in a selective way – an approach that proved to be useful
in such a diverse situation.
In this study, mobility analysis was performed regarding geographical and thematic
mobility. Overall, however, this analysis was rather limited, as not enough detailed
information was available. From the analysis, however, some interesting patterns
emerged. Regarding geographical mobility, a clear separation between the German and
the French / Italian speaking part of Switzerland could be identified. As Figure 14
illustrates, the only link between the German speaking universities on the right side of
the figure and the French (Geneva, Lausanne, FribourgFR, Neuchâtel) and Italian
(Lugano) speaking universities on the left side consists in two professors teaching at the
university of Lugano who earned their PhD in Germany. Also visible in this figure is
the strong presence of professors with German origins in universities in the German
speaking part of Switzerland.
Figure 23: Geographical mobility of professors in Communication sciences in
Switzerland: place of doctorate place of employment
This picture of separation between the French and the German speaking universities
emerged in a similar way also from the analysis of publication lists – at the moment of
data collection, no common communication channel existed, and publications in the
local language were frequent.
94
Regarding thematic mobility, the study has shown that Communication sciences are an
interdisciplinary field: only around one third of the professors for whom this
information was available have a doctorate in this field. Others have background in
fields such as philosophy, psychology, sociology or engineering; they have turned to the
field of communication studies only after their doctorate, and many of them are still
connected to other fields as well, as is visible in their publication lists.
In this study, we were interested in a look at a whole field, thus not at individual
pathways of researchers. For this purposes, CVs retrieved from the Internet showed to
be a valuable source. For a more detailed analysis interested also in individual pathways
of researchers and in factors influencing these pathways, however, it seems that it
would be necessary to ask researchers directly to provide extended CVs and complete
publication lists, or at least a selection of the most important publications.
4.3.6. Conclusions: Possible future developments and recommendations
This report has shown that today the analysis of researchers’ CVs in Switzerland still
entails a lot of manual information retrieving and coding, as no central databases exist
and, where available, information is often heterogeneous and not necessarily complete.
However, in the answers of the representatives of research services or similar units at
the higher education institutions, a slight tendency towards more organised information
as well as interest in CV databases is visible: Several of the contacted persons answered
that the idea of having a university-wide database including also CV information would
be interesting, but that it does not (yet) exist. In some places, new databases are about to
be set up.
At the national level, there is currently no standardised database. Since March 2007,
applications to the Swiss National Science Foundation are done through an online tool
(mySNF), but CV information is uploaded as PDFs. In terms of studies on researchers’
careers, it would be interesting if, in a future version, CV information would be
standardised and included in a database format. This would require a higher effort of the
researchers when introducing their first application, but as already today the system
works with individual account in future applications it would only be necessary to insert
additional, new information.
From the point of view of researchers interested in studying career paths of researchers,
it would be interesting if such databases could contain information that is enough
standardised to be used for further analysis without too much manual processing of the
data. For example, when asking for information on grants, instead of leaving only one
field to fill in all information, single pieces of information such as the granting
institution, starting and ending date or amount could be asked separately. Not only for
CV information, but also for publication lists it would be interesting for further research
to dispose of such separated information.
To have such a database would not only be interesting for researchers interested in
career paths, but also for the institutions creating the databases themselves, thus for the
universities or the SNSF. This information could be used for evaluation purposes, but
95
also in order to get a picture of the population of researchers – a university for example
could use aggregated information in order to show its international orientation, or the
SNSF could more easily get information on the whole population of applicants.
As for the time being, however, no such databases exist, analyses of researcher’s
mobility can be done based on what is already available on the Internet, or by asking
researchers directly for their CVs. As the above example of the analysis of the field of
communication sciences in Switzerland has shown, when designing such a study it is
useful to be aware of the limits of the available data. Also, it proved to be helpful to
clearly define, based on the available data, the aims of the study before starting the
coding procedure, which allows limiting this rather time consuming procedure to the
necessary.
References
Goastellec G., Leresche J.-P., Moeschler O., Nicolay, A. 2007. Les transformations du
marché académique suisse. Evaluation du programme Professeurs boursiers
FNS. Lausanne: Bern: SNSF.
Lepori B., Probst C. 2008. Using Curriculum Vitae for Mapping Scientific Fields. A
small-scale experience for Swiss Communication Sciences. Research Evaluation,
forthcoming.
SNSF Swiss National Science Foundation 2006. Study commissioned by the Swiss
National Science Foundation on Gender and Research Funding (GEFO). Bern
96
5. List of Project Participants
France (OST):
Isabelle Recotillet (irecotillet@free.fr, contact person)
Patrick Werquin
Israel (Technion - SNI):
Zipi Buchnick
Yair Even-Zohar
Daphne Getz (daphne@sni.technion.ac.il, contact person)
Vered Segal
Bella Zalmanovich
Norway (NIFUSTEP):
Anders Ekeland (anders.ekeland@nifustep.no, contact person)
Portugal (DINAMIA):
Emilia Araujo
Sofia Bento
Margarida Fontes (margarida.fontes@ineti.pt, contact person)
Luisa Henriques
Leonel Duarte dos Santos
Spain (UAM-URJC):
Inés Andujar
Carolina Cañibano (carolina.canibano@urjc.es, contact person, project co-ordinator)
Félix F. Muñoz
Francisco Javier Otamendi
Carmen de Pablos
Switzerland (UNISI):
Benedetto Lepori (benedetto.lepori@unisi.ch, contact person)
Carole Probst
United Kingdom (SPRU):
Ana Fernandez-Zubieta
Aldo Geuna (a.geuna@sussex.ac.uk, contact person)
International collaborators:
Barry Bozeman, University of Georgia
María Guillerina D’Onofrio, Universidad de Buenos Aires
Monica Gaughan, University of Georgia
Tim Turpin, University of Western Sydney
Richard Woolley, University of Western Sydney
97
6. Project outcome and diffusion activities: publications and
presentations
Articles:
Research Evaluation Special Issue:
The articles below have been published in the Research Evaluation Special Issue on the
use of CVs in research evaluation (June 2009, Vol. 18, n.2) edited by Carolina
Cañibano and Barry Bozeman.
Cañibano, C. and B. Bozeman (2009) Curriculum vitae method in science policy and
research evaluation: the state of the art. Research Evaluation 18(2): 86-94
Cañibano, C., J. Otamendi and I. Andújar. An assessment of selection processes among
candidates for public research grants: the case of the Ramón y Cajal Programme
in Spain. Research Evaluation 18(2): 153-161
Lepori, B. and C. Probst (2009) Using curricula vitae for mapping scientific fields: a
small-scale experience for Swiss communication sciences Research Evaluation
18(2): 125-134
Zubieta, A. F. (2009) Recognition and weak ties: Is there a positive effect of
postdoctoral position on academic performance and career development?
Research Evaluation 18(2): 105-115
Other articles:
Cañibano, C., J. Otamendi and I. Andújar (2008) Measuring and assessing researcher
mobility from CV analysis: the case of the Ramon y Cajal programme in Spain.
Research Evaluation, 17(1): 17- 31
Working papers:
Crespi, G., D'Este, P., Fontana, R. and Geuna, A. (2009) The impact of academic
patenting on university research and its transfer. ICER, Working Paper.
Submitted papers under review:
Cañibano, C., J. Otamendi and F. Solís (2009) Investigación y Movilidad: análisis de las
estancias en centros extranjeros de los investigadores andaluces. Paper submitted
to the Revista Española de Documentación Científica.
Conference papers:
Cañibano, C., J. Otamendi and F. Solís (2009) International temporary mobility of
Andalusian researchers: a study based on electronic CV data. Annual Meeting of
the Society for Social Studies of Science, Washington, DC, October 28 to
November 1 (accepted).
Fontes, M. and A. Pirralha (2009) Assessing Scientific Mobility Dynamics and Impact:
drawing on the potential of electronic CV databases, 2009 Annual Meeting of
the Society for Social Studies of Science, Washington, DC, October 28 to
November 1 (accepted).
98
Pirralha, A., M. Fontes and J. Assis (2009) Assessing Scientific Mobility Dynamics and
Impact on Knowledge Networks: drawing on the potential of electronic CV
databases, 12th International Conference on Technology Policy and Innovation,
Porto, 13 and 14 July 2009 (accepted)
Pirralha, A., M. Fontes and J. Assis (2009) “Assessing Scientific Mobility Dynamics
and Impacts: the case of mobility during the PhD”, 9th Conference of the
European Sociological Association, 2-5 September 2009, Lisbon (accepted).
Zubieta, A. F., (2008) The recognition of weak ties: the impact of postdoctoral
fellowships on academic performance. International Sociological Association
(ISA) Forum, Barcelona.
Zubieta, A. F., (2008) Mobility of human resources in the UK R&D system. Tracing the
effect of researchers’ mobility on productivity through CVs analysis. II PRIME
Indicators Conference, Oslo.
Zubieta, A. F. (2009) Tracing researchers' mobility. Sectoral job mobility, academic
performance and career development. Triple Helix Conference, Glasgow
Oral presentations:
Cañibano, C. (2008) El análisis de currículum vitae aplicado al estudio del capital
humano científico y técnico en España. Seminario, Instituto de Gestión de la
Innovación y el Conocimiento (INGENIO), CSIC – UPV, Valencia, December
5th 2008.
Cotta, D., S. Bento and M. Fontes (2009) The mobility of Portuguese researchers and
knowledge circulation: methods and preliminary results, 2nd NONIUS Seminar
- Controversies, challenges and policy: Portugal and the European knowledge
triangle, 29 May 2009, Lisboa.
Fontes, M., A. Piralha and J. Assis (2008) "Building Mobility Indicators from the
Electronic CV Database "Plataforma DeGóis”: Some methodological insights",
PRIME Annual Conference "Research and Innovation practice, policy and
theory – changing interactions", 15-16 December 2008, Aix en Provence.
Pirralha, A., M. Fontes and J. Assis (2009) “Building mobility indicators: defining
scientific mobility through electronic CV analysis”, 2nd NONIUS Seminar -
Controversies, challenges and policy: Portugal and the European knowledge
triangle, 29 May 2009, Lisboa.
Pirralha, A., M. Fontes and J. Assis (2009) Metodologias de análise da mobilidade: uso
de curricula, Workshop Mobilidade dos Investigadores e Circulação de
Conhecimento, 30 Maio 2009, Lisboa.
PhD Theses:
Andújar, Inés “Scientific and technical capabilities building process: the role of
researchers’ mobility and networks. A theoretical and empirical assessment
based on the case of Spain”. URJC, Spain. (Ongoing thesis)
Probst, Carole “Understanding the doctorate in Swiss communication sciences”. UNISI,
Switzerland. Defended on May 2009
99
ANNEX 1: Project Seminar Agendas
Building new indicators for researchers’ careers and mobility based on
electronic curriculum vitae: EURO-CV
International workshop
Oslo, May 26th-27th, 2008
Venue: NIFUSTEP. Wergelandsveien 7. Room “Styrerommet", 6th floor.
Monday 26th
12.00 - 13.00 Lunch: NIFUSTEP Cantina
13.00-13.30 Introduction: presentation of participants, general objectives of the project and
specific objectives of the workshop. Carolina Cañibano (Universidad Rey Juan
Carlos)
Session 1: National experiences: availability, characteristics and accessibility of
electronic CV databases and registers
13.30-14.00 Israel: Daphne Getz (Technion-SNI)
14.00-14.30 Norway: Anders Ekeland (NIFUSTEP)
14.30-15.00 Discussion
15.00 – 15.30 Coffee break
15.30-16.00 Switzerland. Carole Probst (UNISI)
16.00-16.30 United Kingdom. Ana Fernández Zubieta (SPRU)
16.30-17.00 Discussion
Tuesday 27th
9.00 – 10.00 France and European Databases (CEDEFOP, Euro-Pass, EURES).
Patrick Werquin (R&W Expertise – OST) – [presentation and discussion]
Session 2: Collecting, coding and analysing CV data
10.00 – 11.00 Scientific mobility and networking in the Asia-Pacific region: towards a
regional CV database. Richard Woolley & Tim Turpin (University of Western
Sydney) – [presentation and discussion]
11.00 – 11.15 Coffee break
11.15 – 12.00 Methodological aspects of CV collection, coding and analysis in the UK. Ana
Fernánfez Zubieta (SPRU) – [presentation and discussion]
12.00 – 13.00 Lunch
13.00 – 13.45 Policy evaluation, mobility and knowledge flows: a CV-data based assessment
of the Ramón y Cajal programme. Javier Otamendi (Universidad Rey Juan
Carlos) – [presentation and discussion]
13.45 – 14.00 Break
14.00 – 14.45 The USA experience: building databases, mapping value and evaluating
research. Monica Gaughan (University of Georgia) – [online presentation and
discussion]
14.45 – 15.00 Coffee break
100
Session 3: Towards national systems of scientific curricular information
15.00 – 15.30 The Portuguese Plataforma DeGois. Emilia Araújo and Sofia Bento
(DIMAMIA)
15.30 – 16.00 The Spanish System of curricular information: recent developments. Carolina
Cañibano (Universidad Rey Juan Carlos)
16.00 – 16.30 The Latin-American experience: developments in the region with special
attention to the Argentine curricular information system. Mª Guillermina
D’Onofrio (Universidad de Buenos Aires).
16.30 – 17.30 General discussion
List of participants:
Family name First name Institution Country E-mail
Andujar Inés URJC Spain ines.andujar@urjc.es
Araujo Emilia DINAMIA Portugal era@ics.uminho.pt
Bento Sofia DINAMIA Portugal bentosofia@oniduo.pt
Buchnik Zipi Technion - Israel zipi@sni.technion.ac.il
SNI
Cañibano Carolina URJC Spain carolina.canibano@urjc.es
D’Onofrio Maria Universidad Argentina mgdonofrio@gmail.com
Guillermina de Buenos
Aires
Ekeland Anders NIFUSTEP Norway anders.ekeland@nifustep.no
Fernandez- Ana SPRU Great Britain A.Fernandez-Zubieta@sussex.ac.uk
Zubieta
Gaughan Monica University of USA gaughan@uga.edu
Georgia
Getz Daphne Technion - Israel daphne@sni.technion.ac.il
SNI
Muñoz Félix F. UAM Spain felix.munoz@uam.es
Otamendi Francisco URJC Spain franciscojavier.otamendi@urjc.es
Javier
de Pablos Carmen URJC Spain carmen.depablos@urjc.es
Probst Carole UNISI Switzerland Carole.probst@lu.unisi.ch
Turpin Tim University of Australia T.Turpin@uws.edu.au
Western
Sydney
Werquin Patrick OST France rw.expertise@free.fr
Woolley Richard University of Australia R.Woolley@uws.edu.au
Western
Sydney
Zalmanovich Bella Technion - Israel bella@sni.technion.ac.il
SNI
101
Building new indicators for researchers’ careers and mobility based on electronic
curriculum vitae: EURO-CV
2nd International workshop
Aix en Provence, December 18th, 2008
Venue: Institut d’Études Politiques, 25 Rue Saporta
09.00-09.15 Introduction: objectives of the workshop. Carolina Cañibano (URJC)
Part 1: National experiences in collecting and analysing CV data from electronic
CV information systems
09.15-10.00 Portugal: "Building mobility indicators from the electronic CV database
"Plataforma deGois": some methodological insights" Margarida Fontes
and Luiza Henriques (DINAMIA)
10.00-10.45 Norway: "Electronic CVs in Norway - publication or career focus?"
Anders Ekeland (NIFUSTEP)
10.45 – 11.15 Coffee break
11.15-12.00 Spain: “Mobility patterns of Andalusian researchers: first results obtained
from SICA” Carolina Cañibano and Javier Otamendi (URJC)
12.00-13.30 Lunch break
Part 2: Experiences in collecting, coding and exploiting non systemic curricular
information
13.30 – 14.15 UK: “Data collection and data analysis in the UK case, problems and
possible solutions.” Aldo Geuna (University of Torino and SPRU) Ana
Fernández Zubieta (IPTS and SPRU)
14.15 - 15.00 Switzerland: “Highly heterogeneous situation." Carole Probst (UNISI)
15.00 – 15.30 Coffee break
15.30 – 16.00 Invited contribution: “Chinese researchers returning home: impacts of
International mobility on research collaboration and scientific
productivity”. Koen Jonkers (CSIC)
16.00 – 17.00 General discussion: reflection on the project’s results, project’s final
report and future research actions.
List of participants:
Family First name Institution Country E-mail
name
Andujar Inés URJC Spain ines.andujar@urjc.es
Cañibano Carolina URJC Spain carolina.canibano@urjc.es
Ekeland Anders NIFUSTEP Norway anders.ekeland@nifustep.no
Fontes Margarida DINAMIA Portugal margarida.fontes@ineti.pt
Fernandez- Ana IPTS & UK, EC A.Fernandez-Zubieta@sussex.ac.uk
Zubieta SPRU
Geuna Aldo University of Italy, UK aldo.geuna@unito.it
Torino &
SPRU
Henriques Luiza Maria IPTS & EC, Portugal luisa.henriques@ec.europa.eu
102
DINAMIA
Jonkers Koen CSIC Spain Koen.Jonkers@EUI.eu
Otamendi Francisco URJC Spain franciscojavier.otamendi@urjc.es
Javier
Probst Carole UNISI Switzerland Carole.probst@lu.unisi.ch
103
ANNEX 2: Database Templates
The following templates provide systematised information on the variables available in
the databases used for exploratory studies.
NORWAY: FRIDA
104
SPAIN: SICA
105
PORTUGAL: DEGOIS
106
UK: SPRU AD-HOC DATABASE
107
