Brenda Reeb
Chris Eagle
Louis King
Chris Hubing (Penn State)
Dan Kiskis (Michigan)
Dana Miller (Miami U of Ohio)
Glenn Donaldson (Ohio State)
Luke Tracy UMich
Michael Janek (MnSCUY)
Mojgan Amini (UCSD)
Noreen Hogan (UofO)
Paul Crum (Penn State)
Rupert Berk (UWash)
Scott Fullerton (Unizin)
Ashish Pandit (UCSD)
Updates on API Working Group
Meetings will be on monthly 2nd Wednesday. Next pres by BYU on API documentation
Spring F2F
Expect survey questionnaire for preferred dates
Implementing Data Governance
University of Rochester
Alan Czaplicki, Deputy to the Provost and Brenda Reeb Information Architect
Problem statement: need for data infrastructure to support research data. No centralized accessible pool of faculty data
Approach; to form data governance and structure.
What was it like to understand the problem and formulate the approach?
- Challenge of understanding data as a shared resource in a decentralized environment.
- Data governance is a thing (i.e., change in perspective) and not just a project.
Is there any way data governance has deterred your work?
- There is a need to balance the need for an established structure and yet there is an urgency to address issues that can't wait for a well established gov structure
Who are the drivers?
- IT folks at U level, IT folks at Medical Center, Provost's office, Being led out of the Provost's office. Question about how leadership will be established. Questions about how to escalate issues to be addressed by the gov group.
- There is a matrix of people with differing levels of knowledge and authority. These categories don't reside cleanly in one place. But data is much too important to be managed in a committee structure. Can go through a matrixed committee process for a few years but may not be the best structure for the longer term.
- Direct lines of authority get blurred when addressing data gov. This can be a challenge to people.
How the need originated
- The focus developed organically. The Associate provost has a broad set of responsibility
Focus is on data quality across domains. Questions about how to determine policies and structures. It seems, though, the focus is on administrative data. Is there attention paid to output from intellectual activities
- Ultimately yes. A longer term goal.
What is focus for next six months
- a clear statement and buy-in about the central importance of well-managed data.
Talk about the role of steward
- have role of steward to address policy and definition.
- have role of owner to field request for access to data (still working on getting that role to hold)
Are you able to leverage drivers?
- new systems may be drivers
- but really the biggest drivers are the needs for better information: assessments, etc.
Yale University
Limor Peer and Louis King.
Will make materials available
Role and current work
Started from an Institute supporting social science research data. Desire in 2010 to make research data publicly available. At the time there was no answer as to how to do that? Questions about workflows, policy, mechanisms. At the time no established institutional repo for research data.
- Data governance
- Research Data Policy
How do you understand data governance today? How has it changed?
- Not changed all that much. Learned a lot along the way.
- Data gov implies coordinated effort involving all the pieces to address needs: involves storage, preservation, sharing.
The biggest problems seem to regarding research output. Louis's group has been attempting to address the general problem from an administrative side hoping it was open to the research side. When the research policy group looked at it it seemed to resonate. Does the framework really apply to Research.
- Framework is needed.
- But policies in place are sometimes in conflict with the needs of research data
- Various tensions are going to be a challenge. Research data policy group feels that it owns the research data (ultimate responsibility). As a result, the researchers act as custodians, but there are greater university responsibilities regarding the data.
- Where there may be a difference between research and admin domains, lies in the rules regarding the release of data.
Drivers include Quality, student access to data, migration to a new ERP. Has led to recognition of a university-wide response.
What have you identified that needs work? How are you constituted?
- differing perspectives: gen counsel, faculty, students
- Immediately identified policy issue regarding access to data – the people authorized to grant access.
- A student focused project (hackathon) led to need to resolve questions about use and authorization to release access. Need to delegate responsibility. Counsel is reviewing policy
- Louis: looking for gov structure that balances opportunity vs risk
Limor urges anyone thinking about data governance to address research data.
How difficult is it to set up institutional data gov policy without an established repository?
- work can continue without established repository, but having an inst repository will help.
How do you deal with informal boundaries/definitions of research data?
- Leaving that to faculty. Generally defines domain as those data supporting published results.