New methods in bibliometric analyses

1
New methods in bibliometric analyses

• Berenika M Webster

2
In search of new bibliometric methods and
techniques

• Measuring impact of a scholarly discipline in the
  periphery (Polish Sociology Citation Index)
• Surname analysis a valid method in assessing
  productivity and impact of groups of scientists
  (women and ethnic minorities)?
• Identifying funders of research

3
What is bibliometrics?

• Bibliometrics as a method of inquiry into volume
  and impact of scholarly activity
• Publication output as a measure of scholarly
  activity of a scholar
• Volume count of publication outputs
• Impact citation counts (for individual papers
  or impact factors for journals)
• ISIs citation indices as an established source
  of data
• Other sources monographs, acknowledgements,
  weblinks

4
PSCI

• Background
• Policy implications for creation of PSCI
• Polish sociology as an example of social
  scientific discipline in a periphery
• Polish sociology as a mirror of social, economic
  and political changes in the country

5
PSCI, or why SSCI does not work for everyone

• National/local importance of social scientific
  research
• Nature of communication in social sciences
• American bias in sociology (casus of Max Weber)
• Structure and coverage of SSCI

6
Main findings

• PSCI is a more sensitive tool allowing for more
  in-depth analysis
• Political and economic changes are reflected in
  productivity of Polish sociologists
• There is little international collaboration
  (despite post-1989 changes
• SSCI shows only the iceberg Polish sociology
  (methodological papers and fashionable topics)
• There are variations in impact of different
  centres (Lodz strong in PSCI, Warsaw in SSCI)
• In both indices, institutes of the Academy of
  Sciences had higher impact than universities
  (teaching informing research?)

7
Surname analysis (1)

• Polish women in science winners amongst losers
• Employment at 40 level, productivity at 32 and
  dropping
• Favourite disciplines include clinical medicine
  and biomedicine little research in physics,
  engineering and maths
• Preference for domestic journals
• Less international cooperation than men
• Better represented in basic than applied research

8
Surname analysis (2)

• Ethnic minorities in UK science
• Publications by ethnic minority authors currently
  constitute over 7 of all UK-only SCI outputs
• In engineering and technology nearly 15 of all
  papers were written by ethnic minority authors
  (fractional count)
• Also in physics, maths, clinical medicine and
  chemistry ethnic minority outputs are greater
  than expected (i.e. they surpass both population
  and employment counts)

9
Surname analysis (3)

• Ethnic minorities in UK science (cont.)
• Chinese and Indians are dominant groups (57 of
  ethnic employment and nearly 70 of ethnic
  outputs)
• Chinese show the biggest share and growth across
  disciplines and concentrate mainly in the areas
  of engineering, mathematics and physics
• Indians dominate in clinical research and
  biomedicine
• Black Africans are behind other groups (0.7 of
  population, 0.6 of academics and 0.1 of outputs)

10
Surname analysis (4)

• Ethnic minorities in UK science (cont.)
• Ethnic minority research is published in lower
  impact journals than that of non-ethnic authors
  they also produce less basic work
• Indian authors produce high impact applied
  research, while Chinese concentrate on basic
  research
• Black African research shows the lowest potential
  impact

11
Identification of funding sources the example
of UK biomedical research 1989-2000

• Nearly 400K bibliographic records were analysed
  as a part of a bigger exercise mapping the
  characteristics of UK biomedical research
• Nearly 9,000 individual funders were identified
  in acknowledgements
• Bradford distribution
• Positive correlation between numbers of funders
  per paper and its impact (up to 6 funders)

12
Percentages of ROD papers acknowledging support
from five main sectors and with no funding
acknowledgement, 1989-2000
UK governments support is acknowledged in
decreasing numbers of papers, while UK PNPs,
foreign and international funders are increasing
their share dramatically (e.g. EC support grew
from around 1,000 papers in late 1980s to over
6,000 ten years later)
13
of funding acknowledgements in 32 subfields
by major funding sectors, 1989-2000