Data Sharing for Demographic Research

1
Data Sharing for Demographic Research

• Center for Family and
• Demographic Research
• Bowling Green State University
• September 13, 2005

2
Public Funding and Data Sharing

• As of October 1, 2003, the NIH requires all
  proposals for over 500,000 annual direct costs
  to include a data sharing plan.
• The final statement on data sharing recognizes
  the rights of investigators to data collection
  "the timely release and sharing" to be no later
  than the acceptance for publication of the main
  findings from the final data set. (NIH Final
  Statement on Sharing Research Data, Feb. 26,
  2003).

3
Data Sharing for Demographic Research

• Partnership project in Cooperative Agreement
• University of Michigan (ICPSR, PSC)
• University of North Carolina (CPC)
• University of Minnesota (MPC)
• NICHD (Demographic Behavioral Sciences Branch)
• Steering Committee of Project PIs
• National Advisory Committee of experts helps set
  project priorities (and includes Wendy Manning)

4
Major Focus

• Data are the central intellectual resource of our
  field
• Resources to preserve share data are limited,
  especially for individual PIs
• DSDR task is to pool our technical capital in
  order to
• Provide resources to data producers to enable
  them to share data
• Provide resources to data users to enable them to
  acquire data

5
The Challenges

• Serve data producers and data users
• Ensure long-term sharing and preservation of data
• Facilitate sophisticated services, such as user
  registration and bibliography management
• Protect research subjects
• NOT GET in the WAY of Anyone!

6
One View of the World of Data
7
DSDRs View of the World of Data
Data User
Data Producer
Data User
DSDR
Data User
Data User
Data User
8
(No Transcript)
9
Resources and Services

• Enhancing the public release of data files
• Providing support for restricted access data
  files
• User support, specialized training, and outreach
• Data archiving actual and virtual
• Research and development in data sharing
• http//www.icpsr.umich.edu/DSDR/

10
Enhancing public release of data Consultation
through the data life cycle

• Data preparation guide
• http//www.icpsr.umich.edu/access/dataprep.pdf
• Individual consultation on data preparation
• Metadata preparation and support
• DDI initiative and training
• Confidentiality review

11
Enhancing public release of data

• Regulating use and release
• User registration identification, download
  counts, updates
• Enhancing data use
• Bibliography development and maintenance
  (including where applicable full text access)
• First and second tier user support
• Accessing and using data

12
Restricted data files

• Confidentiality review and recommendations
• Part of larger ICPSR initiative on best practices
  
• http//www.icpsr.umich.edu/HSP/
• Model restricted use contracts and
  recommendations
• http//www.icpsr.umich.edu/DSDR/rduc/
• Restricted data file management
• Data enclave services

13
User training, support, outreach

• User tools user guides and model user guides
• Metadata training
• ICPSR summer courses
• 2005 co-sponsored with Child Care Archive
• 2006 potential to cosponsor with Survey Research
  Center on survey design and metadata integration
• Propose localized training

14
Archiving options

• Preservation copies
• Public release
• Restricted and enclave release
• Delayed dissemination
• Virtual archive

15
Why Virtual Archiving?

• Allows us to Stay out of the way of the
  research and translational goals of data
  producers but still provide services
• Searchable union catalog entry
• Bibliography
• User tracking
• Virtual gateway
• Turnkey web pages are on their way

16
Research and development our research partnership

• Longitudinal and time series metadata standards
  (ICPSR)
• Web-based data access tools (MPC-UM)
• Data display and visualization for disclosure
  risk (CPC-UNC)
• Spatial data and disclosure risk (CPC-UNC)

17
Before I Conclude Something Completely
Different

• New at ICPSR, UNC, Harvard, Roper Center
• Data Preservation Alliance for the Social
  Sciences seeks to locate Social Science Data
  at-risk of loss from long-term preservation
• http//www.icpsr.umich.edu/DATAPASS/
• Were eager to hear about your data and
  suggestions for classic data to acquire

18
To Conclude

• Were here to help -- data producers, data users,
  and Center staff who work with them
• We want to hear about your projects and ways that
  we can work with you
• Its best if we start early, because sharing data
  is really something that requires work throughout
  the data life cycle
• We have flexible tools and approaches