Transcript PAA on Scientific Data and Information

PAA on Scientific Data and
Information
Roberta Balstad
Chair, PAA Panel
Outline
Data and Information in Science—Then and
Now
Today’s Challenges
The PAA on Scientific Data and Information
Major Findings and Recommendations
2
The Role of Data in
Science—Then and Now
Clinical Sciences
Laboratory Sciences
Observational Sciences
Social Sciences
Modeling and Simulations
3
Access to Information in
Science—Then and Now
Publications:
Editorial process and peer review
Timing of publication
Ownership/copyright
Cost of publication
Libraries:
Preservation and circulation of records
Costs and space
4
The Part Played by
Scientists in D&I—Then
and Now
Scientists’ role in data collection (e.g., they
generally collected their own research
data—and often built the necessary
instruments to obtain data)
Data ownership (e.g., data belonged to the
scientist who collected them)
5
Causes of These
Transformations
Advances in computational power
Increase in sensor capability and complexity
New software for data collection and visualization
Expansion of data and information storage,
dissemination, and access over time and space
New economic and legal basis for ownership of D&I
New analytic methodologies (e.g., models, data
mining, secondary analysis of data, including meta
analysis)
6
Today’s Challenges
Scientists increasingly rely on data collected and
managed by others
Data collection, analysis, and management have
become separate activities with differing
professional requirements and training
The cost of data collection, instrumentation, and
management is is rising disproportionately to other
research expenses
7
More Challenges
Access to D&I and ownership of D&I may be subject to
costly commercial and legal restrictions
The growing emphasis on interdisciplinary research
requires more extensive metadata and
interoperability of databases, hardware, and software
The growth in analyses of change over time requires
careful preservation of today’s databases for use by
future generations of scientists
8
Still More Challenges
Who is ultimately responsible for the D&I on which
science is based?
Who pays??
9
ICSU’s PAA on Scientific
D&I
• International PAA Panel appointed by CSPR in
2003
• Mandate of panel was:
To propose strategic priorities for ICSU over
next 5 to 10 years
To review D&I capabilities in the ICSU family
To advise national and international agencies
• Panel made over 50 recommendations related to
D&I production, management, access and
dissemination
10
Members of the PAA
Panel
Roberta Balstad, chair
Jean Bonnin
Marc H. Brodsky
Liu Chuang
Carlos A. Correa
Norihisa Doi
Ray Harris
Andrew M. Kaniki
Vitaly Nechitailenko
Pierre L.-J. Ritchie
T.B. Rajashekar
11
Key
Stakeholders/Audience
for Report
•
•
•
•
•
•
Scientific community
Science funding bodies (national and international)
Research and educational institutions
Intergovernmental organizations
Commercial data producers and publishers
The media, decision makers, and policy makers
12
Findings and
Recommendations
Data and information constitute a continuum ranging
from raw data to published papers and data
products, such as maps
Scientists now have less control over data collection
and management than when they collected their
own data
Professional data managers are needed to ensure
that today’s scientific data are available for future
generations
13
More Findings and
Recommendations
ICSU should develop D&I strategies for periods of
decades to centuries
New financial models for current and future data and
information management and preservation are
needed
There is insufficient communication and coordination
in data and information activities across fields and
countries
Scientists must be involved in setting data and
information policies, priorities, and practices
14
Recommendation on D&I
Access
Consistent D&I access policies across nations can
benefit both science and public policy
Panel endorsed the ICSU Policy:
Open access to data
Universal and equitable access to publications
15
Overarching
Recommendations of the
PAA
ICSU should assume a leadership role internationally
in identifying and addressing scientific data and
information policy issues
ICSU should develop a long-term, coordinated
framework for D&I policies, practices, and
infrastructure that will operate over decades to
centuries
ICSU should promote cross disciplinary and cross
national collaboration and capacity building by
holding an international Scientific Data and
Information Forum (SciDIF) for data and
16
information stakeholders
SciDIF—and Beyond
ICSU should establish an ad hoc committee to
organize SciDIF and oversee the implementation of
the institutional and policy recommendations of this
report
SciDIF should include all D&I stakeholders and
should be a forum for discussions of broad issues
and needs of current and future scientific research
17