Preserving Research Data The Canadian Experience

1
Preserving Research Data The Canadian Experience

• Charles Humphrey
• University of Alberta
• February 2005

2
Outline

• Two national consultations in Canada
• National Data Archive Consultation (NDAC),
  October 2000 to June 2002
• National Consultation on Access to Scientific
  Research Data (NCASRD), June 2004 to spring 2005
• Findings from these consultations
• Future directions

3
Support Received

• This presentation draws upon research that was
  supported by the Social Sciences and Humanities
  Research Council of Canada (Grant No.
  421-2000-0011 421-2000-0017) and upon the work
  of the National Data Archive Consultation and the
  National Consultation on Access to Scientific
  Research Data.

4
National Data Archive Consultation

• Investigation on behalf of the National Archives
  and the Social Sciences and Humanities Research
  Council
• Two phases
• Year 1 Demonstrate the need for national data
  archiving services
• Year 2 Recommend one or more models to provide
  these services

5
The NDAC Phase I

• The definition of research data employed by the
  NDAC consists of three parts
• outputs of the research process that exist
  between raw research materials and published
  results
• digital information structured through
  methodology for the purpose of producing new
  knowledge
• digital information produced by researchers and
  of interest to researchers.

6
National Data Archive Consultation. Phase One
Needs Assessment Report. May 2001.
7
The Risk Level for Research Data in Canada

• Three studies were conducted in conjunction with
  the NDAC that provided evidence about the level
  to which Canadian research data are at risk.

8
The Risk Level for Research Data in Canada

• A gap-analysis of existing mandates and practices
  of national institutions
• A follow-up study to an investigation first
  conducted twenty years ago by the now defunct
  Machine-Readable Archives and
• A survey of researchers receiving a standard
  research grant from the SSHRC between 1998 and
  2000.

9
Gap Analysis

• The mandates and practices of Canadian
  institutions with responsibilities for the
  preservation of heritage were examined to
  determine the types of digital objects that are
  currently protected.

10
Gap Analysis

• Findings
• The vast majority of academic and non-academic
  research data fall outside the current
  interpretation and execution of the mandates of
  the National Library and National Archives (now
  the Library Archives of Canada).
• No other Canadian institution has a national
  mandate or the resources to address the current
  level of need for preserving research data.

11
Revisiting the MRA Study

• An administrative investigation twenty years ago
  identified a population of 150 SSHRC-funded
  studies utilizing research data.
• Twenty years later, can the data from any of
  these research projects be located?

12
Revisiting the MRA Study

• Findings
• Data from 3 out of 110 studies could be found
  without contacting the original principal
  investigators directly for further details.
• The 3 studies for which data were found were all
  deposited in the United States with the
  Inter-university Consortium for Political and
  Social Research (ICPSR).

13
Revisiting the MRA Study

• Conclusion
• The risk of data loss is very high without an
  institution with the specific mandate to preserve
  research data.

14
A Survey of SSHRC-funded Researchers

• Researchers who received a standard research
  grant from the SSHRC between 1998 and 2000 were
  asked about their plans to preserve the data from
  their projects.
• Only seven percent said that they had deposited
  the data from their funded project, while another
  18 percent said they intend to deposit the data.

15
A Survey of SSHRC-funded Researchers

• When asked to identify where they had or would
  deposit data, almost all named a source that is
  not an archive.
• They mistakenly listed university library data
  services, the Web, and a Statistics Canada
  Research Data Centre.
• A couple of researchers indicated that they would
  deposit the data from their projects if they only
  knew where and how to do this.

16
A Survey of SSHRC-funded Researchers

• Conclusion
• Without a recognized institution responsible for
  preserving research data, researchers do not know
  where or how to archive the data from their
  research, even if they would like to see the data
  preserved.

17
A Survey of SSHRC-funded Researchers

• Conclusion
• For the vast majority of researchers in this
  study, archiving data is an unknown activity in
  conducting research.

18
Size of the Problem

• We know that research data are at risk, but how
  big of a problem is this?
• The survey of researchers who received a standard
  research grant from the SSHRC provides evidence
  that around 550 out of every 1,000 projects
  results in the creation or use of data files
  and/or databases.

19
Size of the Problem

• This is just the tip of the iceberg!
• There is no estimate for other SSHRC-funded
  projects, other granting agencies, or other
  agencies and departments creating research data.

20
Does It Matter?

• This question has been asked in other countries
  with answers that apply equally in Canada.
• Protecting the financial investment in data
• Stewardship and custodial responsibilities
• Legal and ethical obligations and
• Knowledge-generation opportunities.

21
The NDAC Phase II

• A primary objective of the second phase was to
  recommend the institutional form that national
  data archiving services should take in Canada.
• First, research was conducted to identify the
  types of existing institutional models for data
  archives.

22
The Results

• A typology of organizational models was developed
  from the results of a survey of 36 international
  organizations in data archiving and data services
  in the social sciences and humanities.
• Three generalized models were identified that
  summarized groupings of the characteristics from
  the survey.

23
The Results

• While no single existing institution is
  necessarily described completely by one of these
  three models, the typology offers a fair summary
  of the current mix of organizations.

24
The Three Models

• The Topical Data Archive
• The Agency-based Data Archive
• The Comprehensive Research Data Archive

25
A Proposed Canadian Model

• Establish by legislative mandate an agency
  reporting to Parliament through the Ministry of
  Industry or Heritage or a combination of both
• Fund centrally through Parliament
• Grant authority to act on behalf of the
  Government of Canada in international
  negotiations related to research data and its
  management standards and practices
• Structure as a network of distributed service
  points with a central service facility

26
A Proposed Canadian Model

• The central facility would be responsible for
  data management, standards development, and data
  preservation
• The service points would be responsible for
  assisting with the deposit of data, accessing
  data, and training and user consultation
• Service points would be located in universities
  and other institutions interested in providing
  access to preserved research data (a model
  similar to the Depository Service Program between
  government publishing and Canadian libraries)

27
A Proposed Canadian Model

• A management board would oversee the operation of
  this National Data Archive Network and consist of
  representatives from the regions in Canada as
  well as various stakeholders that manage, use,
  and produce research data
• Furthermore, this agency would enter into formal
  co-operative working relationships with other
  national institutions, such as the Library and
  Archives of Canada and Statistics Canada.

28
National Consultation on Access to Scientific
Research Data

• A Task Force of experts was assembled to organize
  a two-day National Forum to investigate issues
  regarding access to research data in Canada and
  to formulate recommendations.
• Experts from the natural and medical sciences
  were engaged to complement the work of the NDAC.
• The Task Force developed a mind-map of ideal
  achievements reached in 2010 as a result of
  improved data access.

29
(No Transcript)
30
The NCASRD National Forum

• A document structured around the main entries of
  this mind map was prepared and distributed to
  an assembly of 70 researchers who attended a
  National Forum held on November 22 23, 2004.
• This body generated its own mind map of
  achievements for the year 2010.

31
(No Transcript)
32
(No Transcript)
33
The NCASRD National Forum

• Working backwards from the ideal achievements,
  sub-groups prepared vignettes of the steps needed
  to be accomplished to reach the end-states
  documented in the mind map. These steps were
  subsequently organized into recommendations.
• A draft report with recommendations arising from
  the discussions at the National Forum has been
  written and circulated to members of the Task
  Force. Look for the final report to be available
  before the summer of 2005.

34
Future Directions

• A field of funded research is needed to study
  issues about the preservation of and access to
  research data.
• Areas for research
• The Data Economy who gets what data, when and
  how
• The Life Cycle of Data the course of data and
  its corresponding metadata from the earliest
  stages of planning through to the secondary uses
  of data

35
Future Directions

• Areas for research
• Metadata Standards the development of standards
  that detail the life course of data
• Preservation Standards the best practices in
  long-term preservation of data
• Data Stewardship the ethics and best practices
  of sharing research data
• Inhibitors to Access the legal and cultural
  barriers to access.

36
References

• Hackett, Yvette. A national research data
  management strategy for Canada the work of the
  National Data Archive Consultation Working
  Group. IASSIST Quarterly, vol. 25, no. 3 (2001),
  13 -16.
• Humphrey, Charles. On the advantages of freely
  accessible data comment letter. Epidemiology,
  vol. 14, no. 3 (2003) 381.
• Humphrey, Charles. Research for building a
  better data community. IASSIST Quarterly, vol.
  25, no. 1 (2001), 21-24.
• Jacobs, James A. and Charles Humphrey.
  Preserving research data. Communications of the
  ACM, vol. 47, no. 9 (September 2004) 27-29.

37
References

• National Data Archive Consultation. Phase One
  Needs Assessment Report. May 2001.http//www.sshr
  c.ca/web/whatsnew/initiatives/da_phase1_e.pdf
• National Data Archive Consultation. Final Report
  Building Infrastructure for Access to and
  Preservation of Research Data. June
  2002.http//www.sshrc.ca/web/whatsnew/initiatives
  /da_finalreport_e.pdf
• National Consultation on Access to Scientific
  Research Data website.http//ncasrd-cnadrs.scit
  ech.gc.ca/about_e.shtml