Author: Anna Sackmann

With the increase in data science across all disciplines, most undergraduates will encounter basic data science concepts and be expected to analyze data at some point during their time at UC Berkeley. To address this growing need, the Library Data Services Program began partnering with the Data Science Discovery Consultants in the Division of Computing, Data Science, and Society (CDSS) on the Introduction to Data Analysis Support workshop series in Fall 2020. The Data Science Discovery Consultants are a group of undergraduates majoring in computer science, math, data science, and related fields who are hired as student employees. They receive training to offer consultation services across a wide range of topics, including Python, R, SQL, and Tableau, and they have existing partnerships with other groups on campus to provide instruction around data as a part of their program. Through the partnership, Data Science Discovery Consultants work with librarians to develop as instructors and gain experience constructing workshops and teaching technical skills. The end result is the creation of a peer-to-peer learning environment for novice undergraduate learners who want to begin working with data. The peer-to-peer learning model lowers the barrier to learning for other undergraduates and enhances motivation and understanding.

The Data Science Discovery Consultants enthusiastically embraced the core values of the Carpentries, through which they empower each other and the audience, collaborate with their community, and create inclusive spaces that welcome and extend empathy and kindness to all learners. In Fall 2022, attendance for the workshop series was opened up to local community college students who may be interested in transferring to UC Berkeley. One of the workshops was taught in Spanish, to provide an environment in which native Spanish speakers could better connect with one another and the content. 

Diego Sotomayor, a former UC Berkeley Library student employee and current Data Science Discovery Consultant, taught the inaugural Introduction to Python in Spanish: Introducción al análisis de datos con Python. Diego comments that: 

“Languages at events are no longer just a necessity but have gone to the next level of being essential to transmit any relevant information to the interested public. There are many people who only speak Spanish or another language other than English and intend to learn new topics through various platforms including workshops. However, because they are limited by only speaking a language that is not very popular, they get stuck in this desire to progress and learn. Implementing the workshop in different languages, not only in English but in Spanish and even others, is important to give the same opportunities and equal resources to people looking for opportunities.”

The UC Berkeley Library and the Division of Computing, Data Science, and Society hope to further provide these offerings for prospective transfer students in Fall 2023. Many thanks to Elliott Smith, Lisa Ngo, Kristina Bush (now at Tufts University), and Misha Coleman in the Library. Anthony Suen is the Library’s staff partner in the Data Science Discovery Program and Kseniya Usovich assists with outreach. 

The Library Data Services Program recently posted about the National Institutes of Health (NIH) Data Management and Sharing Policy and how it will affect UC Berkeley researchers. Please read more about the new policy on this post.

Here are a list of FAQs about the new policy. Please contact the UC Berkeley Library Data Services Program (librarydataservices@berkeley.edu) with questions.

How is UC Berkeley  responding to this policy?

The Library Data Services Program is collaborating with the Research Data Management Program to provide guidance and documentation to ensure compliance with the NIH policy.

What is considered “scientific data” for the purposes of this plan?

The final NIH Policy defines Scientific Data as: “The recorded factual material commonly accepted in the scientific community as of sufficient quality to validate and replicate research findings, regardless of whether the data are used to support scholarly publications. Scientific data do not include laboratory notebooks, preliminary analyses, completed case report forms, drafts of scientific papers, plans for future research, peer reviews, communications with colleagues, or physical objects, such as laboratory specimens.” The NIH states that “the final DMS Policy is designed to increase the sharing of scientific data, regardless of whether a publication is produced…Data that do not form the basis of a publication produced during the award period should be shared by the end of the award period.”

What is included in a Data Management and Sharing Plan?

In these max two-page documents, researchers will describe their:

• Data type(s)
• Related tools, software, and/or code
• Standards
• Data preservation, access, and associated timelines
• Access, distribution, or reuse considerations
• Oversight of data management and sharing

Read more about Data Management Plans and see sample language

Can I make my data available upon request?

NIH strongly prefers that scientific data be shared and preserved through repositories or, for datasets up to 2GB, through PubMed Central-deposited supplemental data files, rather than kept by a researcher and provided upon request.

How will the plans be assessed?

NIH program staff will assess the DMS plans but peer reviewers may comment on the proposed budget for data management and sharing.

What data repository should I use?

NIH encourages the use of established repositories. To select the best repository for your data consider the following:

• Is there a specific NIH repository named in the funding announcement?
• Is there a data repository specific to your discipline?
• If not, is there a general data repository you can use?
  • For UC Berkeley researchers, consider Dryad data repository 

To learn more, read the NIH guidance on selecting a data repository.

What is a standard? What standards are relevant to my research?

A standard specifies how exactly data and related materials should be stored, organized, and described. In the context of research data, the term typically refers to the use of specific and well-defined formats, schemas, vocabularies, and ontologies in the description and organization of data. However, for researchers within a community where more formal standards have not been well established, it can also be interpreted more broadly to refer to the adoption of the same (or similar) data management-related activities, conventions, or strategies by different researchers and across different projects.

When do I need to make my data available?

NIH encourages scientific data to be shared as soon as possible, and no later than time of an associated publication or end of the performance period, whichever comes first.

What data management and sharing costs can I include in my grant?

Allowable costs can include:

• data curation and developing documentation (formatting data, de-identifying data, preparing metadata, curating data for a data repository)
• data management considerations (unique and specialized information infrastructure necessary to provide local management and preservation before depositing in a repository)
• preserving data in data repositories (data deposit fees)

Read more about allowable costs.

Guidance for data management and sharing costs on NIH budget requests

What happens if I do not comply with the NIH policy or make my data available as described in the DMS policy?

NIH Program Staff will be monitoring compliance with the policy during the funding period. “Noncompliance with Plans may result in the NIH ICO adding special Terms and Conditions of Award or terminating the award. If award recipients are not compliant with Plans at the end of the award, noncompliance may be factored into future funding decisions.”

I work with sensitive topics/populations – how do I protect my participants’ privacy?

NIH strongly encourages researchers who work with sensitive topics and/or populations to address data sharing in the Informed Consent process. See the UC Berkeley Human Research Protection Program’s Informed Consent page, which includes guidelines and appropriate form templates.

Researchers should pay special attention to their de-identification process to ensure that all identifying information has been fully removed. Researchers should consider depositing their data in restricted access repositories that require data use agreements and research plans in order to access the data. Contact librarydataservices@berkeley.edu if you would like guidance on selecting restricted access repositories.

Please view the UCSF’s resources on data de-identification and sharing de-identified data for additional guidance. 

Do specific NIH Institutes and Centers (ICs) have additional policies or recommendations? 

Yes, NIH ICs may have additional requirements or recommendations. Please identify the institute or center using this table to learn more about requirements.

Supplemental information from the NIH:

Responsible Management and Sharing of American Indian/Alaska Native Participant Data

Protecting Privacy When Sharing Human Research Participant Data

Many thanks to Ariel Deardorff at the UCSF Library for allowing us to adapt their list of Frequently Asked Questions and thank you to UC Berkeley’s Elliott Smith, Michael Sholinbeck, and Erin Foster for all of their expertise and contributions. 

On January 25, 2023, the National Institutes of Health (NIH) will implement a new Data Management and Sharing Policy. The Library Data Services Program and Research Data Management Program have several resources to help you adapt to the new policy, including extensive guidance and suggested language for writing a plan. Additionally, the DMPTool, which is free for UC Berkeley users and supported by the Library Data Services Program, walks grant applicants through the plan requirements. We will be offering four drop-in workshops designed for researchers this fall. Please register using for the workshops using the links below:

• Monday, September 12th: 2:10-3:00 PM (this session will be recorded)
• Monday, October 10th: 2:10-3:00 PM
• Monday, November 14th: 2:10-3:00 PM
• Monday, December 12th: 2:10-3:00 PM

The NIH is a leader in implementing data management plans and was the first federal granting agency to do so with their 2003 NIH Sharing Policy. In the years since, the agency developed two genomic data sharing policies (2008 and 2014), and addressed data sharing in clinical trials in 2016. The new data sharing policy builds on their existing data management requirements and is broadly sweeping with the goal to maximize the “…sharing of scientific data generated from NIH-funded or conducted research, with justified limitations or exceptions.” 

The new policy will apply to all research funded (in whole or in part) by the NIH that produces scientific data. It will apply to grant applications submitted on or after January 25, 2023. The NIH defines scientific data as: “the recorded factual material commonly accepted in the scientific community as of sufficient quality to validate and replicate research findings, regardless of whether the data are used to support scholarly publications…” Please note that the NIH definition of scientific data does NOT include the following: laboratory notebooks, preliminary analyses, completed case report forms, drafts of scientific papers, plans for future research, peer reviews, communications with colleagues, or physical objects, such as laboratory specimens.

There are a few aspects that set the new NIH DMSP apart from current policy:

• Plans will outline how data and metadata will be managed over the course of the project and which of these data will be shared.
• Grant applicants will need to include details about software or other tools that were used to analyze the data.
• If generating data derived from human participants, plans will need to outline how confidentiality, privacy, and rights of those individuals will be preserved.
• Plans must include a selected repository or repositories where the data will be preserved along with a timeline for sharing the data (either as soon as possible, no later than the time of an associated publication, or at the end of performance period if there is no associated publication). 
• Updates to plans over the course of the project will be reviewed by the NIH ICO (institute, center, or office) during regular reporting intervals.
• Data management costs may be added to the grant budget including data curation and developing documentation (e.g., formatting data, de-identifying data); data management considerations (e.g., unique and specialized information infrastructure necessary to provide local management and preservation before depositing in a repository); preserving data in data repositories (e.g., data deposit fees).
• Compliance with plans will be measured during the funding period at regular reporting intervals.

Please see our list of frequently asked questions. For additional information about the policy, please see the NIH page on Data Management and Sharing Policy. 

Many thanks to Elliott Smith, Michael Sholinbeck, and Erin Foster for all of their expertise and contributions!

If you have any questions or need additional support, please email librarydataservices@berkeley.edu.