The Secrets to Successful Research Data Services
The Research Data Access and Preservation (RDAP) Summit 2015 started Wednesday, April 22nd with a room packed full of data and research services librarians, specialists, and managers, data curators, archivists, and professors, just to name a few. RDAP is unique conference in that it’s smaller (about 100 attendees) with a single track of sessions for all attendees. RDAP brings together data professionals focused on providing access to and preserving research data. The sessions are specifically designed to reach across disciplines to find what challenges occur regardless of context and what differentiates one context from another.
Keynote: Building Research Data Services at Research Universities
Dr. Claudia Neuhauser, Director of the University of Minnesota Informatics Institute (UMII)
Research is about creating knowledge collaboratively. UMII was created to accelerate this creation and exchange of knowledge by focusing on people and processes in an effort to anticipate research data needs throughout the entire data lifecycle. This is a complex process that involves extracting data in myriad formats that must be harmonized and cleaned if it is ever to be used effectively.
“80% of work in any data project is cleaning the data.” – D.J. Patil, U.S. Chief Data Scientist
Furthermore, it requires providing support at every step in the data lifecycle, whether by identifying appropriate tools for analyzing data or working with data on behalf of researchers. Dr. Neuhauser considers big data to be any data set that a researcher considers to be too big for them to effectively work with, regardless of size. Moreover, a significant foundational aspect of research services is policy development. For this reason, UMII developed a policy that defined research data in such a way that it encompassed all individuals involved in the data lifecycle, regardless of whether they were a researcher or metadata specialist. Other policies defined ownership and stewardship of data, archiving, retention, security, storage, and transfer. Policies in tandem with robust research data services reveals both the short- and long-term data management needs, enables meeting of regulatory obligations, and supports evaluation of data management solutions.
Panel: You’re in Good Company: Unifying Campus Research Data Services
Cynthia Hudson-Vitale, Digital Data Outreach Librarian, Washington University
Brianna Marshall, Digital Curation Coordinator, University of Wisconsin-Madison
Amy Nurnberger, Research Data Manager at Columbia University
The keynote segued nicely into the first panel in which the panelists from three different academic institutions shared their experiences of integrating research data services on their respective campuses. Panelists agreed that research data issues cannot be addressed solely by libraries, but require strategic partnerships.
“Data is the new oil.” –Ann Winblad
Data may be the new oil, but the process for curating data must account for every step in the data curation lifecycle. Planning must support growth, security, and integrity of data. Data must be managed so that it is useful and renewable. Data must be analyzed to identify solutions to real world problems. And finally, data must be preserved in order to support ongoing research.
All of the panelists have had to advocate for research data services, stepping outside of their comfort zones to get answers that will move processes forward and raise the visibility and campus awareness of research data services. The process for establishing research data services requires building infrastructure to support research activities, creating bridges or points of connections between stakeholders, and getting the right tools for the job. Policies play an integral part in all processes by formalizing previously informal arrangements so that people can be held accountable for the success of research data management. But most important is the ability to make decisions despite the chaos that may surround you. One must maintain awareness of opportunities as they arise so that they can be leveraged to enhance data research services.
“A New York minute is sixty seconds of possibility.” –Charles Lewis Tiffany
Panel: Research Data Integration in the Purdue Libraries
Amy Barton, Metadata Specialist, Purdue University
Pete E. Pascuzzi, Molecular Biosciences Information Specialist, Purdue University
Line Pouchard, Computational Science Information Specialist, Purdue University
Tao Zhang, Digital User Experience Specialist, Purdue University
Purdue has well-developed research data services and the attendant culture. Services start at the planning phase of grants and research projects, offering data management plan development and metadata standard recommendations. Additionally, the Purdue University Research Repository (PURR) provides project collaboration and project management features, as well as digital object identifiers (DOIs) for research data.
Describing the potential of data reuse and citation helps to motivate students and faculty to participate in research data services. User experience then plays a large role in designing user-friendly research data services that focus on usability, engagement, and satisfaction and seek to attain the greatest level of acceptance by users. When the user is effectively engaged with research data services and tools, it’s more likely that the necessary data curation steps will be taken.
Panel: Developing Data Literacy Programs: Working with Faculty, Graduate Students, and Undergraduates
Jake Carlson, Research Data Services Manager, University of Michigan
Lisa Johnston, Data Management Librarian, University of Minnesota
Amy Koshoffer, Science Informationist, University of Cincinnati
Megan Sapp Nelson, Engineering Information Specialist, Purdue University
Research data management is typically not taught as part of any core curriculums, especially if you’re not involved in data or information sciences. So, how do research data professionals go about improving research data literacy?
At the University of Michigan, they developed twelve digital information literacy competencies by which to gauge the roles librarians could play in supporting research data management, to create communities of practice, to align current research data practices with research needs, and to understand research needs depending on research contexts.
“Use data terminology that offers the most traction in getting your message across.” –Jake Carlson
The University of Minnesota started offering a data management course to increase digital literacy among students, providing guidance on organization and documentation, data access and ownership, data sharing and reuse, and preservation techniques. This course combines asynchronous online and in-person data management coursework.
Evaluation and assessment of research data needs and users is critical to providing relevant, useful solutions. In teasing out the needs, policies and procedures can be developed to overcome high turnover of students supporting faculty, standardizing data management skills, and providing training and support to achieve a baseline research data management literacy. The University of Cincinnati even went so far as to incorporate research data management literacy requirements in their strategic plan, accentuating it as a vital skill to be acquired by faculty and students alike.
Jumpstarting a Research Data Program with Limited Resources
Mary Molinaro, Director of the Research Data Center, University of Kentucky
Raise awareness of research data services before building a team or creating a community.
Research Data Services at the University of Colorado Boulder
Shelley Knuth, Senior Research Data Specialist, University of Colorado Boulder
Incentivize best practices by creating research data management competitions for seed funds.
Virginia Tech University Libraries’ Data Service Pilot with the College of Natural Resources and Environment
Natsuko Nicholls, Research Data Consultant, Virginia Tech
Workflows must allow for creativity and spontaneity.
University Data Policies and Library Data Services
Abigail Goben, Assistant Information Services Librarian, University of Illinois-Chicago
Create separate data policies rather than lumping them in with IT to ensure they address the entire data lifecycle.
Again and again over the course of the day, people repeatedly spoke to the necessity of creating a culture that supports effective research data management. And because research is so incredibly collaborative, it is also a distributed activity. Building infrastructure to support distributed research data management must happen in tandem with policy development to clarify roles at every point of the data lifecycle, processes involved in research data management, and the needs of researchers.