Research Data Management @Harvard
- 1. Research Data Management @Harvard Mercè Crosas, Ph.D. @mercecrosas Chief Data Science andTechnology Officer Institute for Quantitative Social Science Harvard University
- 2. “Good data management is not a goal in itself” Wilkinson et al.,The FAIR Guiding Principles for Scientific Data Management and Stewardship, Nature Scientific Data, 2016
- 3. “Good data management is not a goal in itself” •Enables continuity of research projects •Facilitates data sharing and re-use •Reduces research and data storage costs •Helps with data reproducibility
- 4. Connecting Computing Resources and Data Management is critical Harvard Data Group (established in 2016 by Office of Vice-Provost of Research; Tahmassian and Crosas) Research Computing Council (established in 2016 by CIO; Margulies, Cuff) Data (Crosas) Access (Yockel) Talent (Adair) HMS Data Management Working Group (established in 2014, grassroots) •Office of Vice-‐Provost of Research •IQSS •HUIT •HMS IT •HU Library •Countway Library •HU Office of Sponsored Programs •HMS Basic Sciences •HMS Sponsored Programs Admin 2014 2017 •Countway Library •HMS IT •HMS Basic Sciences •HMS Sponsored Programs Admin •Harvard Chan BioinformaHcs •HMS Research CompuHng •HMS Academic and Research Integrity •HUIT •FAS Research CompuHng •HMS Research CompuHng •HBS Research CompuHng •IQSS •HU Library
- 5. Harvard Data Group has concreteTasks • Build a website for research data management @Harvard, coordinating with all existing resources (Spring-Summer 2017) • Create a research data management training module, with custom modules for various research domains (2017-2018) • Data User Agreements sub-group to coordinate DUA tracking and workflows, as part of data management support (2017) • More in the future
- 6. A Single Entry to Data Management Avoids Confusion • For researchers, not for librarians, archivists, or trainers • Cite scholarly work, evidence- based studies • Concise; point to other resources as needed • “good enough data management”: • what you need to know • what Harvard can offer • other resources you can use
- 7. RDM @Harvard will link to HMS Data Management and Library sites
- 8. Data ManagementTraining offered for Medical School & School of Public Health • Organized by the HMS Data Management Working Group • Based on the data lifecycle for biomedical research • Has been offered a few times in 2016/2017 • Will be combined with Harvard wide training
- 9. Extension of Harvard Dataverse Curation Services • Led by Sonia Barbosa (IQSS) • 6 month pilot program with Harvard librarians • Offers extended curation services to Harvard affiliates (and all users, when possible) • Evaluating cost-based model • Plus, office hours once a week
- 10. Data Management Support is not Sufficient Research Computing & Security Support Data Science Support Data Management Support LayersofSupport
- 11. DataFest 2017 Brings Data Science BasicTraining to researchers and staff
- 12. More technology integration and ease-of-use, less training and support
- 13. Data Repositories can help Integrate the Data Lifecycle
- 14. Dataverse is an open-source platform for building any type of data repository, including institutional repositories. A growing community of developers and users http://dataverse.org Agriculture*data** Repository*in* Fudan,*China* Data*from*20*Universi>es* Public*data*repository* Science*Consor>um*
- 15. An Integrated Data Management and Computing Solution E Lab Notebooks, Instruments, Surveys, … Storage and Computing Journals Data Repository Assign Security Level, DUA DOI, Metadata, DUA, Restrictions Link citations DOI, metadata, and DUA are assigned after data collection; Data repository enables data-centric computing Track Provenance Metadata
- 16. Machine-readable, FAIR Data Management Plans can help track data management
- 17. THANKS! Mercè Crosas @mercecrosas mercecrosas.com With contributions from Caroline Shamu, Radhika Khetani, and Sonia Barbosa In summary: • Coordinate, coordinate, coordinate (across groups) • Integrate, integrate, integrate (across technologies)