Acrl march2015 final
Download as PPTX, PDF5 likes1,465 views
This document summarizes roles for libraries in providing research data management services. It describes data services at the University of Oregon Library including consultations, education workshops, and developing data management web pages. It discusses support for documentation provided by the University of Idaho Library through instruction sessions, research consultations, and emphasizing good documentation practices. It outlines data management trainings provided by Oregon Health & Science University Library including workshops with researchers, individual consultations, and developing new data services.
![Spreadsheet for Help with
Organizing
Research
Project:
[Name of research
project]
Name: [Your name]
Dates:
[when you'll be
conducting your
research, e.g. 7/14-
1/15]
Project Data
Folder:
[e.g.
dissertation_coldfusion
_data]
Research
Process/Method
/ Data Source
Collection
Dates Storage Format
Original
Format
Working
Format Access Format
Preservation
Format(s)
File Naming
Convention
Folder /
Convention Versioning Strategy
Storage
Location Who can help?
Access
restrictions?
Who
needs
access?
Software /
Tools Required
Metadata
Schema Notes](https://image.slidesharecdn.com/acrlmarch2015final-150323122617-conversion-gate01/85/Acrl-march2015-final-20-320.jpg)
Acrl march2015 final
- 1. Roles for Libraries in Providing Research Data Management Services Nicole Vasilevsky, Oregon Health & Science University Victoria Mitchell, University of Oregon Jeremy Kenyon, University of Idaho
- 2. Nicole Vasilevsky Project Manager, Biocurator and Ontologist, Ontology Development Group, OHSU Victoria Mitchell Social Science Data & Government Documents Librarian, University of Oregon Jeremy Kenyon Research Librarian, University of Idaho Library
- 3. 1 | Data services at UO Library 2 | UI support for documentation 3 | OHSU data management trainings
- 4. Do you have experience in data management training?
- 5. Why do our patrons need to know about data management?
- 7. Why? Researcher Perspective Version control Track processes for reproducibility Quality Control Stay Organized Save Time and Stress Avoid Data Loss Format data for reuse (by self, team, or others) Document for own recollection, accountability, reuse
- 10. At the UO Libraries Data Services
- 11. The UO Environment • No campus-wide research data policy • Library leading on research data management and preservation • Collaborating with campus IT, Research Services
- 12. The UO Environment • Digital Scholarship Center • Open Access Publishing • Digital Collections • Institutional Repository • Interactive Media Development • Data Services • Science Data Services Librarian • Social Science Data Librarian
- 13. Services • Data Management Plans – Consultation and review
- 14. Data Management Web Pages
- 15. Services • Consultations with faculty • Special projects – Southern Voting Project
- 16. Education • Workshops • Presentations in classes and new faculty orientations • 1-credit course in research data management for grad students
- 17. Graduate Seminar in Data Management • 2 iterations so far • 1st: Spring 2013 – 1 credit course, LIB 407/507 • Made it available to upper-division undergrads; none signed up • 2nd Spring 2014 – 1 credit course, LIB 607
- 18. Graduate Seminar in Data Management Based course around creation of a DMP for a funding agency • Students registering for the course were strongly encouraged to have a research project already in mind or underway • Also used, in part and with modification, the education modules created by DataONE
- 19. • Natural disaster • Facilities infrastructure failure • Storage failure • Server hardware/software failure • Application software failure • External dependencies (e.g. PKI failure) • Format obsolescence • Legal encumbrance • Human error • Malicious attack by human or automated agents • Loss of staffing competencies • Loss of institutional commitment • Loss of financial stability • Changes in user expectations and requirements Data Loss CCimagebySharynMorrowonFlickr CCimagebymomboleumonFlickr Slide adapted from DataONE Education Module: Why Data Management. DataOne. Retrieved March 21, 2013
- 20. Spreadsheet for Help with Organizing Research Project: [Name of research project] Name: [Your name] Dates: [when you'll be conducting your research, e.g. 7/14- 1/15] Project Data Folder: [e.g. dissertation_coldfusion _data] Research Process/Method / Data Source Collection Dates Storage Format Original Format Working Format Access Format Preservation Format(s) File Naming Convention Folder / Convention Versioning Strategy Storage Location Who can help? Access restrictions? Who needs access? Software / Tools Required Metadata Schema Notes
- 21. LIB 607 v.3 • Changed to Data Management for the Social Sciences (and Digital Humanities) • Less emphasis on DMP per funder requirements • More time to address issues specific to the social sciences and humanities
- 22. @ the University of Idaho Library Research Data Services Credit: University of Idaho Creative Services
- 23. University of Idaho Characteristics: • Public, comprehensive, land-grant university • Strong emphasis on agriculture, environmental science, engineering • Recent emphasis on developing research data and research cyberinfrastructure, including library research data services, INSIDE Idaho, the geospatial data repository, and NKN, a multi-disciplinary institutional data repository
- 24. How do we move from this?
- 25. To this?
- 26. To this?
- 27. Research Data Services at the U-Idaho Library Appointments & Consultations Northwest Knowledge Network (institutional data repository) Embedded Services (Buy-outs of librarian time)Tool & Technology Support: IQ-Station, ESRI Products, DMPTool, Metadata editors Website: Data Management Best Practices Guide Instruction & Workshops Many modes of service Raise awareness of research data management & our services Create a culture of documentation Transform thinking across disciplines about data distribution & publishing
- 28. Focus: creating a culture of documentation FISH502 “One-shot” Instruction Session - Class participants: fisheries biology and statistics graduate students - Exercise: 1) review the following spreadsheet 2) identify the information needed to re-use this dataset
- 29. Focus: creating a culture of documentation Research consultation: environmental modelling Post-doc from a multi-institutional project was primary contact for several teams Consultation on metadata was made towards the end of project Producing 6 discrete collections of data as netCDF (format required by funder) Repository required ISO 19115 XML metadata for describing whole collections
- 30. Focus: creating a culture of documentation Challenges: Understanding the standard Attribute Conventions for Dataset Discovery ISO 19115-2 Codelists and controlled vocabularies Rules for free-text fields what does a good title look like? Placement of content should variables be listed in keywords, title, or description? Responsibilities who should create XML files – the researcher or us?
- 31. Focus: creating a culture of documentation Re-use and comprehension of data requires good documentation Researchers often have idiosyncratic and localized, i.e. customized, documentation practices Content standards are often not well-known among researchers Disciplinary content standards are necessary for enabling advanced modes of data access Library services must emphasize documentation
- 32. Future Directions Fienberg, S.E. et al. (1985). Sharing Research Data. Washington, D.C: National Academies Press. http://www.nap.edu/catalog/2033/sharing- research-data
- 33. at Oregon Health & Science University Research Data Management Efforts
- 34. What would you do with $1k today to make research communication better that doesn’t involve building another tool?
- 35. 1| Workshops with the library 2| Individual consultations
- 39. Your Data: Gummy Bear Raw Data Bounces Amplitude Color 15 4 blue 43 3 red 58 9 green 75 82 purple Materials: • Haribo Gummi Bears Sugar Free, 5 lb bag, Amazon.com (UPC: 422384500110) • SpringOMatic 3000 (ICanPickleThat, Portland, OR) http://laughingsquid.com/the-anatomy-of-a-gummy- bear-by-jason-freeny/
- 40. Figure 1. A) Gummy skeleton with belly button annotated with red arrow B) Springiness by sample color. Methods Section: Haribo Gummi Bears (Sugar Free) were purchased from Amazon.com (UPC: 422384500110). Gummy bears were placed in the SpringOMatic 3000 (ICanPickleThat, Portland OR) according to the manufactures instructions. The Gummy Anatomy (Jason Freeny) image was cropped in PPT (Microsoft) and annotate to highlight the bellybutton. Gummy Bear Final Figure 0 2 4 6 8 10 12 14 16 blue red green purple Springiness(bounces/length) Sample Color A B Figure legends/metadat a Manipulating images Attribution Metadata about research resources
- 42. Group 1: Gummy Bear Final Data 0 2 4 6 8 10 12 14 16 blue red green purple 4 3 9 82 15 43 58 75 Springiness (Bounces/Amplitude) 15 4 blue 43 3 red 58 9 green 75 82 purple Methods: A schematic of a Gummi Bear was cropped to indicate where the belly button is located (Fig. 1). At this point, raw experimental data showing the bounce, amplitude, and color were analyzed and the springiness calculated for each color of bear. This was accomplished by dividing the bounce by the amplitude and plotting this against bear color. Fig. 1 Belly button of Haribo Sugar Free Gummi Bear What is missing? A.Image manipulation B. Attribution C. Figure Legends D.Metadata about resources
- 43. Figure 1. A) Gummy skeleton with belly button annotated with red arrow B) Springiness by sample color. Methods Section: Haribo Gummi Bears (Sugar Free) were purchased from Amazon.com (UPC: 422384500110). Gummy bears were placed in the SpringOMatic 3000 (ICanPickleThat, Portland OR) according to the manufactures instructions. Group 2: Gummy Bear Final Data 0 2 4 6 8 10 12 14 16 blue red green purple Springiness(bounces/length) Sample Color A B What is missing? A.Image manipulation B. Attribution C. Figure Legends D.Metadata about resources
- 44. Figure 2: Schematic depiction of Haribo Gummi Bear umbilical skeletal anatomy. Methods & Materials Gummi Bears were obtained through Amazon in 3 kg bags. Lot and temperature during transport data were not made available. Bears were housed in a plastic bowl in accordance with IACUC policy and national standards for gummi bear care. They were housed at room temperature on a natural light cycle. Food and water were provided ad libitum (consumption was not monitored) Each bear was sampled only once to reduce costs Group 3: Gummy Bear Final Data What is missing? A.Image manipulation B. Attribution C. Figure Legends D.Metadata about resources
- 45. Belly Button 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 blue red green purple Springiness(bounces/amplitude) Gummy Bear Color (a) (b) Fig. 1. (a) schematic of the anatomy of a gummy bear (adapted from 1). (b) springiness of bear by color using spring-o-matic. Methods: Insert the sample of interest, specifically a colored gummy bear (Haribo, Japan). Position the probe above the sample. Press "Tickle" and the SpringOMatic (ICanPickleThat, Portland) will poke the belly button a standard depth of 1 cm. Record the number of bounces and the amplitude of the largest bounce in cm. From these values, the springiness can be calculated (bounce/amplitude). What is missing? A.Image manipulation B. Attribution C. Figure Legends D.Metadata about resources Group 4: Gummy Bear Final Data
- 46. GUMMY BEARS TAUGHT US… • People see the same data very differently • “Detailed” means different things… • Metadata?!? • File management is difficult • Workflow Vasilevsky N; Wirz J, Champieux R, Hannon T, Laraway B Banerjee K, Shaffer C, and Haendel M. Lions, Tigers, and Gummi Bears: Springing Towards Effective Engagement with Research Data Management (2014). Scholar Archive. Paper 3571.
- 47. CONSULTATIONS Researcher + 2-3 from Data Stewardship Team
- 48. Researchers DO need assistance: Finding and choosing data standards File versioning Applying metadata to facilitate data sharing “Gummi Bear” themed data management exercise resonated well with students Lack of awareness of services and expertise offered by the Library Conclusions
- 49. OHSU New Directions OHSU Library is developing data services for researchers BD2K educational grants in collaboration with DMICE www.ohsu.edu/xd/education/library/data
- 50. Acknowledgements OHSU Melissa Haendel Robin Champieux Jackie Wirz Kyle Banerjee Bryan Laraway Chris Shaffer Kaiser Todd Hannon UO Brian Westra Karen Estlund Cathy Flynn- Purvis John Russell Idaho Bruce Godfrey Nancy Sprague Lynn Baird Greg Gollberg Luke Sheneman Steven Daley-Laursen
- 51. Contact us Nicole Vasilevsky vasilevs@ohsu.edu @N_Vasilevsky Thank you Victoria Mitchell vmitch@uoregon.edu @VictoriaStap Jeremy Kenyon jkenyon@uidaho.edu @jr_kenyon
Editor's Notes
- #2: Added some placeholder logos
- #4: Outline
- #5: Ask the audience to raise their hands
- #6: Why | Funding agencies are creating mandates to develop data management and sharing plans, and additionally, there is increased focus on reproducibility of science and other disciplines that stems from a need for improved data management.
- #9: Victoria is going to add a different slide with more examples.
- #10: As professionals in curation, organization and classification of information, librarians are well poised to assist researchers by providing data management services and training.
- #13: Soc sci data librarian: More recently created (partial) position
- #16: Consultations with faculty about data produced by their research, their needs for collecting, managing, etc., data; depositing data in our repository Also, Northwest Indian Language Institute – Endangered Languages
- #17: E.g., Office of Research and Innovation – workshop for new faculty on grant-writing for NSF and NIH – give us a little time.
- #20: EXAMPLE of slide borrowed from DataONE
- #21: Use as in-class exercise; students keep adding information as course progresses
- #37: At the DMOH, we discussed topics including scholarly attribution, data sharing, managing your scholarly footprint. At the DMOH, we had researchers attend from various career levels, from grad students to post-docs, to core lab directors to PIs.
- #39: While the research at OHSU is primarily focused on biomedical health research, the specific research projects vary quite greatly, from bench science, to clinical research, across topics such as cancer biology or biomedical engineering. We wanted to come up with an interactive exercise, where we could demonstrate some of the importance of data management skills at each step, but centered around a topic that was either not too specific to someone’s field or too distant from their field. We chose a topics that was fictional and playful- we asked them to pretend they were doing a study that assessed the “springiness’ of a gummi bear
- #40: These are the materials that we gave the students
- #41: This is best viewed as the “slide show”, so you can see the animations. I wanted to point out what the ideal results would be, and some of the key attributes we wanted them to take away.
- #43: I am going to show them all the results from each group, then ask them to raise their hands to answer ‘what is missing’. For example, for this group, the is missing all of the options.
- #44: This group is missing attribution of the image.
- #45: I left off the graph here because I was running out of room.
- #48: At the DMOH and Data Wrangling session, we recruited individual researchers to schedule individual consultations with us. We had X # sign up and X # follow through with consultations. We found, even with the incentive of the gift card, it was difficult to recruit researchers to participate in the consultations.