Poster Session

The posters available here were presented at the 6th Annual University of Massachusetts and New England Area Librarian e-Science Symposium on April 9, 2014. The e-Science Symposium encourages libraries to collaborate and support e-science initiatives at their institutions. Featuring presentations by nationally recognized leaders in the e-Science arena and posters that showcase innovative library projects supporting scientific research, the symposium is an educational opportunity for librarians to learn about e-Science resources and current initiatives. The symposium also provides a forum where librarians can discuss new library roles for engaging research communities and supporting networked science.

Subscribe to RSS Feed

2014
Wednesday, April 9th
12:45 PM

An Assessment of Doctoral Biomedical Student Research Data Management Needs

Kate Thornhill, Simmons College GSLIS
Lisa A. Palmer, University of Massachusetts Medical School

12:45 PM

Objective

This analysis explores specific institutional repository (IR) data management needs of the University’s biomedical sciences doctoral students. Awareness, intentions, attitudes, and concerns about depositing, sharing and publishing supplemental ETD (electronic thesis and dissertation) research data into the library’s institutional repository eScholarship@UMMS were explored.

Methods

A data needs assessment survey focused around the Digital Curation Centre’s lifecycle model and National Science Foundation’s requirements for data management was sent to 470 students via a listserv. Information gathered from the survey and digital repository literature aided in the construction of an overarching student data curation profile and criteria for repository functionality to meet the needs of both researchers and the repository manager.

Results

Eighty-two biomedical PhD students responded to the data needs survey, a response rate of 17.4%. 69.5% were unaware that they had the option to deposit their research data sets into the IR. File format of data sets varied greatly but most common were TIFF, PDF, and JPG. 25.6% of respondents did not know the average size of their data sets. A network shared drive was the most common means of storing data (75.0%) but many used multiple methods. 96.0% reported using a metadata data entry standard developed by their lab. 13.9% stated they would not be willing to share data sets openly or publicly.

Conclusion

Responses from the survey and interviews suggest that an IR needs to be flexible to accommodate the research data needs of biomedical PhD students. Functionality to handle various file types, large files, and embargos is required. Education and outreach by library staff about the IR, data documentation, data sharing, and many facets of research data management would be useful. A broader environmental scan and further research are required to evaluate repository functionality in light of the needs of both researchers and the repository manager.

Describing Data Repositories

Jennifer M. Eustis, University of Connecticut - Storrs
Carolyn Mills, University of Connecticut - Storrs

12:45 PM

With the rise of eScience, subject liaisons must become familiar with disciplinary data repositories to better serve their clientele. Research data can often be deposited in one or more repositories. For researchers who are not well informed or work in fields that have yet to develop a data repository existing lists such as DataBib, Registry of Research Data Repositories or OpenDOAR provide a combined list of up to 2000 data repositories but little information about each one. Subject liaisons at the University of Connecticut Libraries can help researchers find appropriate data repositories for data submission and discovery. However, with such a large listing, how do subject liaisons evaluate repositories in their disciplines? To support our subject liaisons better evaluate data repositories and to give them more confidence to help their faculty in eScience, we created the “Describe Your Data Repository” survey.

This survey ultimately has two aims: aid subject librarians become familiar with data repositories in their subjects and encourage their involvement in eScience; offer researchers a thorough and easily understood review of relevant data repositories. It does this by asking 37 questions in 5 sections about data and repositories, including questions about functionality and policies. It can take anywhere from one to two hours to complete depending on the clarity and completeness of information at the repository website. Our poster would like to cover the following points:

  • Conceptualization and design of this survey
  • Getting subject liaisons on board with the survey as a way of advocating for their increased roles in eScience
  • Helping subject liaisons understanding and completing the survey: tips and scheduled group work times
  • Lessons learned
  • Results of the survey and where all this information will go
  • Next steps

Developing a Data Management Curriculum for Graduate Students in the Natural Resources

Sarah J. Wright, Cornell University
Camille Andrews, Cornell University

12:45 PM

Objective

E-science has created a need for new skill sets for graduate students managing, working with and curating their research data. How can librarians help teach data management skills?

Methods

Librarians at Albert R. Mann Library collaborated with faculty in the Department of Natural Resources to determine educational priorities and deliver a one-credit special topics course in Spring 2013. Several assessment methods were used in order to determine the success of the instruction, including in-class active and collaborative learning exercises, post-class one-minute reflections, and a final survey.

Results

Data Management, Data organization, Data quality and Documentation, Data analysis and Visualization, Metadata, and Data Sharing were identified as the top educational priorities, and the course included sessions on each of these topics. Enrollment – almost 30 students of all experience levels – was much higher than expected. The course was co-taught by the librarian and the faculty, with responsibilities divided fairly evenly. Post-class and in-class assessments allowed us to respond quickly to correct misunderstandings or provide additional information when necessary, and the final survey indicated an increase in the students’ self-confidence in their skills and ability regarding every data management topic addressed in the course. As of Spring 2014, the course has been accepted by the curriculum committee, and will become a regular offering of the department.

Conclusions

Graduate students and faculty recognize a need for training in managing research data. Librarians can help to answer this need, providing real educational value for the institutions they serve.

Helping Biomedical Researchers Gain the Credit They Deserve

Varsha Khodiyar, F1000Research
Cesar A. Berrios-Otero, F1000Research
Rebecca N. Lawrence, F1000Research

12:45 PM

In an era of large-scale biomedical research, generating and sharing datasets in an open manner is an important, but non-trivial task. However, researchers are subject to the ‘publish or perish’ culture, where career progression and tenure is highly dependent on publishing papers in peer reviewed journals with high impact factors. Standard journals often have limited space available for each paper, thus much of the scientific literature has little data associated with each article.

In addition, the publication of a dataset is rarely considered as having as high an impact compared with a data analysis paper. There are also numerous technical obstacles in making datasets truly accessible. These issues combine to create a scientific culture where sharing and publishing data ends up low on a researchers’ list of priorities. However, open data can be beneficial to scientific progress in several ways; for example enabling data to be verified1 or the testing of novel hypotheses that were unforeseen at the time of data generation2.

F1000Research is working with funders and institutions to begin addressing some of these challenges. We have implemented several initiatives to provide methods and tools to capture the production of scientific data, and to establish this as an important output of research activity in itself.

References

Simonsohn U, 2013."Just Posting It" works, leads to new retraction in Psychology. Data Colada [blog] 17th September [Accessed: 20 Jan 2014]

Chappell, P. R. and Lorrey, A. M., 2013. Identifying New Zealand, Southeast Australia, and Southwest Pacific historical weather data sources using Ian Nicholson's Log of Logs. Geoscience Data Journal (http://dx.doi.org/10.1002/gdj3.1) [Early View (Online Version of Record published before inclusion in an issue)]

Implementing a Graduate-level Data Information Literacy Curriculum at Oregon State University: Approach, Outcomes and Lessons Learned

Amanda L. Whitmire, Oregon State University

12:45 PM

Purpose

This poster examines the development and inaugural offering of a credit-bearing graduate-level course in data information literacy (DIL). The purpose of the course was to enable students to acquire foundational knowledge and skills in DIL that would support long-term habits in planning, management, preservation and sharing of research data.

Setting/Participants

Oregon State University Libraries partnered with the Graduate School to facilitate the process of establishing a new credit-bearing course in the OSU catalog. The course is open to graduate students from all disciplines without any prerequisites, and is taught by the Libraries’ Data Management Specialist.

Brief Description

This poster describes the development and implementation of a new course in data information literacy for graduate students. Curricular materials were drawn from the New England Collaborative Data Management Curriculum (NECDMC), the DataONE Education Modules, and the MANTRA online course materials. The pedagogical approach for the course combined outcomes-centered course design with active learning techniques. The approach was predicated on the idea that getting the students actively engaged with the content and techniques of data management would be more effective than lecture alone. This approach was also driven by the practical reality that while the course content was necessarily discipline agnostic, successful learning outcomes depended on the students’ ability to apply the material to discipline specific standards of practice (e.g. metadata and data documentation, data sharing formats and methods, etc.). The goal was that after taking the course, the students would successfully incorporate data management best practices into their daily workflow. Such behavioral change takes self-reflection ("How does this material relate to me?") and practice, both of which I strove to offer during class meetings.

The 2-credit class met twice per week for 50 minutes. For each meeting a lesson plan was created that included anticipated timing, learning outcomes, lecture content, teaching strategies, student products, the DIL core competency addressed and assessment approach, if any. Lesson plans also included the associated readings and homework assignment, if any. Not every specific learning outcome had an assessment component, but closely related outcomes were assessed together. Periodic course assessment was also performed through anonymous student surveys, with the objective of gauging course efficacy and quality, and to obtain suggested modifications or improvements. The midterm assignment for the course was a scaled back Data Curation Profile, and the final exam assignment was to create a data management plan for their research project.

Results/Outcome

The inaugural course had an official enrollment of 11 students, including one faculty member enrolled for credit and two as non-credit auditors. I also had an additional faculty member “sitting in” on lectures who did not complete any of the outside assignments. The disciplinary range of the students was broad: six students from the College of Public Health and Human Sciences, two from the College of Forestry, and one each from the Colleges of Veterinary Medicine, Science, and Agriculture. Aside from the faculty members, student degree paths ranged from non-thesis master’s to Ph.D., with some of the students having a very well defined research project already planned and others much less so.

With all of the variability in student disciplinary background and experience, in-class learning activities and the homework, midterm and final exam assignments were relied upon to facilitate application of the generalized course content to their individual, discipline-specific circumstances. The content and quality of the students’ assignments demonstrated that this approach was successful.

Conclusions

The course is not yet complete at the time of this abstract submission, but the poster will summarize the during- and post-class surveys that were used to determine the success and impact of the instruction. Suggested modifications for next year’s course will also be discussed.

Librarians Working with Institutional Stakeholders to Support Researchers Writing Data Management Plans: Piloting a Research Data Management Curriculum

Andrew T. Creamer, University of Massachusetts Medical School
Donna Kafel, University of Massachusetts Medical School
Elaine R. Martin, University of Massachusetts Medical School
Regina Raboin, Tufts University

12:45 PM

Biomedical researchers face multiple challenges in maintaining proper stewardship of their research data. The collaborative nature of team science and funders’ requirements for data management plans have revealed a broad need for adherence to consistent research data management (RDM) best practices. UMMS and partnering libraries have developed the New England Collaborative Data Management Curriculum (NECDMC), a case-based RDM curriculum to support researchers’ writing of data management plans. This poster presents ways that NECDMC is being used by a sample of pilot sites and initial feedback from researcher participants.

Scanning the Data Environment at the University of Massachusetts Medical School

Rebecca Reznik-Zellen, University of Massachusetts Medical School
Sally A. Gore, University of Massachusetts Medical School
Lisa A. Palmer, University of Massachusetts Medical School
Andrew T. Creamer, University of Massachusetts Medical School
Donna Kafel, University of Massachusetts Medical School
Len L. Levin, University of Massachusetts Medical School
Elaine R. Martin, University of Massachusetts Medical School

12:45 PM

Objective

Environmental scanning constitutes a primary mode of organizational learning” (Choo 1999). In a step toward active development of research data support services for its community, the Lamar Soutter Library at the University of Massachusetts Medical School has undertaken extensive environmental scanning to better understand the strengths, weaknesses, opportunities and challenges of an academic biomedical institution with respect to research data. Given the variety of potential data services that an academic library may deploy, the information gathered from these activities will identify and prioritize new library activities.

Method

Environmental scanning activities include a survey of student’s experiences and attitudes with research data management; faculty and administrator interviews (via the DuraSpace 2014 eScience Institute program); and the identification of existing local services and policy documents related to research data. Results from these activities are analyzed by the Library Data Services Advisory Group and the eScience Institute working group to plot a formal roadmap for library-based data services.

Results

Students, faculty, administrators, and existing documentation together reveal a variety of attitudes, assumptions, and avenues for the handling of research data on campus. They identify potential activities where the library might play a role, some expected and some unexpected.

Conclusion

Information gathered during environmental scanning activities at the University of Massachusetts Medical School informs the development and prioritization of library-based research data support services.

The Unified Astronomy Thesaurus: Bringing Together People, Terms, and Organizations to Classify Infinity... and Beyond!

Katie Frey, Harvard-Smithsonian Center for Astrophysics

12:45 PM

Objective

To create a Unified Astronomy Thesaurus as an open, interoperable, and community supported thesaurus. The UAT will bring together existing Astronomy and Astrophysics vocabularies into a single, freely-available, online thesaurus, available to publishers, authors, and everyone else with an interest in classifying astronomical content.

Methods

Because astronomy and astrophysics are fast-moving fields, terms that are common today, such as ‘exoplanets’ and ‘dark energy’, were relatively unknown 10 years ago. In order to keep up with the ever-changing field of astronomy, the UAT will solicit suggestions and recommendations from the astronomy community at large. A librarian will be in place to direct suggestions to an appropriate subject matter expert and facilitate discussion. The subject matter experts will evaluate the suggestion and recommend revisions to the UAT. These revisions are reported to the librarian for implementation versioned releases.

Results

We anticipate the result of this project to be a continually evolving controlled vocabulary, which will be used as a common language across publishers and platforms to connect articles, data sets, and astronomical objects. We also hope this unified vocabulary will inspire a new range of cross-silo data sharing tools.

Conclusion

Thus far the UAT has received considerable support from the American Institute of Physics, Institute of Physics, the American Astronomical Society, and the Astrophysics Data System. We have begun collecting a list of volunteers to edit the UAT and are continuing our evaluating of tools and resources to facilitate the continuous evolution of terminology in astronomy and astrophysics.

Using Zebrafish to Do Good: Scientific Data Management

Julie Goldman, Simmons College

12:45 PM

E-science expands the scope of science library practices and promotes, among medical and graduate/undergraduate science students, the preservation of scientific data in relevant repositories/archives. Case studies can be used as a tool to educate and teach both medical and library science graduate/undergraduate students about the preservation of scientific data. In the medical sciences, researchers must submit a data management plan in order to secure funding for a research project. Therefore, they must understand the best practices that should be followed in their different disciplines.

This case study addresses institutional research in a biomedical neuroscience laboratory at a prestigious research university, conducting experiments with live animals during a long term research project, and also the use of paper lab notebooks. These concepts are included in the narrative of the research story, and then pulled out to model the seven parts of a data management plan. It identifies user requirements and interface design elements for a system that can host student research data; outlines curriculum frameworks and learning needs for research data management instruction that can be delivered through a variety of methods; and presents a communication plan to inform others about the curriculum planning process and results. By using this case study as an E-science tool to help students, they will understand data management principles and challenges in the context of familiar research settings, the benefits of preserving scientific data, and also how these practices will lead to a more homogenous research future.

What to Do about Data: An Overview of Guidelines and Policies for Dataset Collection Development

Sarah Young, Cornell University

12:45 PM

Objective

Datasets are increasingly emerging as a ‘new currency’ in collection development. While purchasing models may in some ways mirror more traditional forms of electronic information, there are many unique considerations in the collection and acquisition of datasets. The purpose of this study is to determine the extent to which academic libraries have formalized dataset collection development policies and to highlight some of the key considerations in the development of such policies. The focus here is on commercially available datasets, rather than datasets produced at home institutions.

Methods

Currently existing dataset collection development policies and guidelines will be gathered from web searches of academic library websites, calls to listservs and personal communications. Based on these existing practices, as well as a brief literature review, key considerations will be identified. Ongoing discussions at our own institution will also inform this work.

Results

Some existing policies have already been identified highlighting several important considerations in the development of dataset collection development policies. Those considerations include cost, individual vs. institutional access, long-term value of the data, storage and preservation, access and discoverability, licensing restrictions, among others.

Conclusions

While several institutions have formalized collection development policies in regards to datasets, this remains a relatively underdeveloped area of collection development. Given the growing importance of datasets as a currency of research, libraries should strive to identify their roles in dataset collection and consider guidelines for selectors, liaisons, and other librarians involved in supporting academic research.