nahenke – Library News

Mar 29 2023

DMPTool Helps Researchers Write Data Management and Sharing Plans

Posted on March 29, 2023March 29, 2023 by Nancy Henke

Because funding agencies like the National Institutes of Health (NIH) and the National Science Foundation (NSF) now require data management and sharing plans as part of grant applications, researchers tasked with writing these documents may feel at a loss about where to start. Luckily, University of Iowa researchers have support in this process. Not only does Research Data Services at the UI Libraries offer workshops, consultations, and web resources to facilitate the writing of data management and sharing plans, we’re also partnering with DMPTool to offer even more assistance for researchers.

DMPTool is a free, open-source web platform designed specifically to help users write data management and sharing plans. Originally created by eight research institutions (including UCLA, UC San Diego, University of Virginia, and the Smithsonian Institution), DMPTool now has over 300 partners – including the University of Iowa.

Here’s how DMPTool can be useful:

It offers step-by-step instructions. DMPTool walks users through the process of writing their plans with a click-through wizard. It has fill-in fields for the basics, like project title, abstract, and contributors, while also guiding users through describing their data types, standards, and preservation strategies. It even offers example wording someone can use as they craft their plan.
It includes funder-specific templates. Since different granting bodies require different elements in their data management and sharing plans, DMPTool offers templates tailor-made to meet the needs of various funders. In addition to heavy hitters like NIH and NSF, DMPTool also has built-in templates for many other agencies – including the National Oceanic and Atmospheric Administration (NOAA), National Endowment for the Humanities (NEH), United States Geological Survey (USGS), United States Department of Agriculture (USDA), and dozens more.
It integrates local help from the University of Iowa Libraries. Because Iowa is a DMPTool partner institution, our researchers have access to some added features – like the ability to submit your plan directly to a data librarian at the Libraries so you can receive feedback on it. When writing your plan you can also see UI-specific tips by clicking the box under “Select Guidance” in the Project Details tab. If you have questions at any point in the process, DMPTool also includes links to Research Data Services’ email and website at the top of every page. Just ensure you set up your account with your UI email address and these services will automatically be integrated into your dashboard.

If you’re ready to try out DMPTool, our website has some tips for getting started. Whether or not you use DMPTool to write your data management and sharing plans, Research Data Services is here to support you. We’re available by email, offer guidance on our website, put on workshops, and do one-on-one consultations as well.

Mar 07 2023

The Power of Persistent Identifiers in Data Sharing

Posted on March 7, 2023 by Nancy Henke

Every researcher has likely seen Digital Object Identifiers (DOIs) attached to the articles they read, cite, and publish. As unique codes that distinguish a specific research article, dataset, or other digital object, DOIs make it much easier to find and cite the work of other scholars. What you may not know, however, is that DOIs are just one type of persistent identifier (PID). PIDs make your research findable now and in the future, reusable for the long term, and offer new insights into your research.

PIDs, sometimes referred to as persistent digital identifiers (PDIs) or globally unique identifiers (GUIDs) take many different forms. You may have seen an Open Researcher and Contributor ID (ORCID) attached to someone’s email signature, or a Research Resource ID (RRID) in a journal article identifying a specific antibody used in the study, or even an International Standard Book Number (ISBN) in a book you’ve read. All of these are PIDs, but don’t let the alphabet soup of initials intimidate you. In essence, all persistent identifiers have two traits that make them powerful.

The first is the part you see and use: the string of letters and/or numbers that uniquely define an entity (like an article, researcher, dataset, or institution). These codes will not change, will not be reused, and are a major reason finding information is so easy with PIDs. The second part of a PID is behind the scenes: the service that locates the resource (or “resolves” it) if its URL changes, ensuring that people who use the PID can always access what they’re looking for. This means, for instance, that even if the URL for an article changes the DOI never will.

Here’s an example: When a researcher deposits their dataset in the University of Iowa’s institutional repository, Iowa Research Online (IRO), a DOI for the dataset is reserved. A data librarian at UI Libraries will then review and curate the data. When the creator of the dataset gives the green light, IRO will register the dataset, its corresponding metadata, and the DOI with an organization called DataCite, activating the DOI and telling DataCite the landing page for the dataset (i.e., the URL that describes the data and provides access to the files). If IRO reorganized its servers in the future and the URL of the landing page changed, IRO would update this “address change” with DataCite. This ensures that the DOI will still point people to the new, correct URL.

The image below demonstrates how the DOI for a dataset resolves to the specific URL in IRO.

Diagram with DOI and corresponding citation on the left, and URL and corresponding landing page on right. — Image adapted from: Nosé, M. (2019). “Practice of research data management in solar-terrestrial physics [PowerPoint Slides]. Institute for Space-Earth Environment Research, Nagoya University. https://slideplayer.com/slide/17406586/

Another important benefit of using persistent identifiers is connection: linking the digital object to everything else that’s associated with it through PIDs, like the people who contributed to the work (through ORCIDs), the places they work (through Research Organization Registry IDs, or RORs), the cell lines they used in the research process (through RRIDs), and articles that use the data (through DOIs).

PIDs thus create a stable network of linked data that makes research outputs more FAIR (Findable, Accessible, Interoperable, and Reusable), helping people find new connections between data, articles, researchers, institutions, and granting bodies. A researcher who finds your dataset might then be curious if you’re the same person who wrote an article on a similar topic they read in a journal. This might lead them to click your ORCID to discover other research you’ve done, pointing them to the DOI of another dataset you’ve published, which they discover they could use for their own project. This discovery was facilitated by PIDs.

The image below offers an example of how PIDs connect these different entities.

You can do a few things to make the most of PIDs and leverage these connections:

Most importantly, deposit your data and code into repositories that support PIDs. This enables these vital connections between your dataset and the people, articles, institutions, and granting bodies associated with it. These valuable links are the reason the National Institutes of Health strongly encourage using PID-friendly repositories in their guidelines for Data Management and Sharing Plans.
Register for an ORCID if you don’t already have one, and use it in your repository deposits, code documentation, CV, personal website, grant applications, and anywhere else you can. It connects others to your research and keeps people from confusing you with another researcher with a similar name. The ORCID system will harvest and connect your research outputs – meaning you don’t have to.
Use RRIDs in your articles if the journal you’re submitting to supports them. The RRID website includes a list of journals that specifically ask for RRIDs (including Cell, Nature, and PLOS One) as well as a list of journals that will include them in articles if the authors add them.

Each of these steps is small and may not seem impactful on its own, but when all researchers do a few small things to enable FAIR data through PIDs, the entire research ecosystem benefits. If you’d like to learn more about repositories like IRO that support PIDs, reach out to Research Data Services in the Scholarly Impact Department of UI Libraries by email or through our website.

Feb 27 2023

Data Curation: Adding Value to Your Dataset

Posted on February 27, 2023 by Nancy Henke

You may already know that the University of Iowa’s institutional repository, Iowa Research Online (IRO), provides both preservation and access to your dataset for the long term. You may not know, however, that Research Data Services in the Scholarly Impact Department at the UI Libraries also offers another key service to researchers depositing their data in IRO: data curation.

Although the term curation (from the Latin “care” or “attention”) might be a practice you associate with historical artifacts or priceless paintings, data curation is a collaborative, value-added process that provides care and attention to a dataset. It helps make data more FAIR (findable, accessible, interoperable, and reusable).

When you deposit your dataset in IRO, a data librarian at the Libraries will work with you to ensure that it is as complete, understandable, and accessible as possible. Data curation is different than peer review; its purpose is to ensure that the data can be found and used, not to judge the scientific methods that went into its creation.

Think of data curation as an investment, and working with a data librarian up front can get you a great return on that investment. It gives you a dataset that’s more valuable to a potential user because it’s easier to find, use, and interpret. This is a service that sets IRO apart from many other repositories, most of which simply don’t have the staff to offer this value-added process.

Depending on the specific dataset, data curation may entail:

checking to ensure all files open properly
reviewing file naming and organization strategies to ensure they’re transparent for future users
identifying proprietary file formats and recommending open alternatives
analyzing documentation, like data dictionaries or README files, to ensure others with knowledge of the discipline can understand them
ensuring tabular data in spreadsheets is clean, organized, and optimized for reuse

Confirming that the dataset has rich and complete metadata attached to it is an essential aspect of data curation. Metadata is something librarians talk about often, and with good reason. It’s the information needed for others to find, understand, and use the data. Different types of metadata – descriptive, administrative, and technical – all add value to your data in different ways, and a data librarian can help ensure the thoroughness of it all.

Descriptive metadata, for instance, is vital for discoverability (i.e., ensuring your data will appear in the results when someone does a relevant search on Google or InfoHawk+, the University of Iowa Libraries’ discovery tool). It includes having a clear and distinct title for the dataset, that all collaborators are named with contact information and ORCIDs (Open Researcher and Contributor IDs) included, and that there is a full abstract.

A data librarian can also help with technical metadata, like specifying the type of software someone needs to open the file, and administrative metadata, such as helping choose the license for the dataset. This information helps a potential user know how to get access to your data and what they can do with it once they do.

If you’d like help with data curation or depositing your data into IRO, Research Data Services is here to help. Contact us by email or visit our website to set up a consultation. If you’re ready to deposit data in IRO, we’ve created a metadata guide and a data deposit guide to walk you through the steps.

Feb 20 2023

Why Iowa Research Online is an Ideal Place for Your Data

Posted on February 20, 2023April 13, 2023 by Nancy Henke

University of Iowa researchers are increasingly taking advantage of the university’s institutional repository, Iowa Research Online (IRO), to house their research and creative works. IRO currently holds nearly 115,000 research outputs from Iowa faculty, staff, and students, and has seen more than 12 million downloads of content since 2009. On top of preserving articles, books, conference proceedings, theses, and dissertations, IRO is also an ideal place for researchers to deposit their datasets and code.

“Research Data Management” by Janneke Staaks CC BY-NC 2.0

IRO provides preservation, access, and curation of your data. Here’s what that means in practice:

Preservation

Your dataset or code will be housed on a secure server for the long term, maintaining it for future use. Despite the perception that digital files never wear out, they can deteriorate. Sometimes this happens due to bit loss – when the binary code that makes up the file degrades as the data is transferred from one place to another – or sometimes because of corrupted files. In other cases, file formats evolve and need to be converted to a different format to enable access and use.

The items in IRO are proactively managed to guard against these situations. Regular fixity checks act as “check-ups” for files to ensure they’re healthy and haven’t changed, multiple copies of the data are archived in different geographic locations in the case of a natural disaster, and corrupted files have self-healing capabilities thanks to the cloud infrastructure that houses them. All of this ensures that the products of your hard work are available now and in the future.

Using open formats for your work also helps with preservation, since these formats – like .sav, .mp3, and .mp4, to name a few – are more likely to remain functional in the future. In fact, using open formats also facilitates another benefit of archiving in IRO, access, since the files don’t require specialized proprietary software to open and use them.

Access

Your dataset will be accessible to researchers all over the world, increasing the reach and impact of your work. This access is made possible by a few key features.

First, all IRO deposits have a metadata record with pertinent information about the dataset – like the title, collaborators, abstract, grant information, dates of data collection, etc. Since all items in the repository are discoverable on Google and are indexed and searchable in InfoHawk+, the University of Iowa Libraries’ discovery tool, robust metadata makes it more likely that your dataset will appear in relevant searches. This is vital for helping others find your work.

Equally important is the stable, persistent URL your dataset will receive. Since the URL won’t change, it eliminates the tedium of identifying and updating broken links on your CV or personal website and makes it easier for you to share your work with others. And if you’re ever curious about the number of views and downloads a dataset receives, the metrics are readily available.

All IRO deposits also receive a digital object identifier (DOI) which makes it easy for others to cite your work when they use it and ensures you get credit when they do. When you deposit your data, you can also link to the DOI of the article or articles where the data is used. Research Data Services in the Scholarly Impact Department at the UI Libraries can even reserve a DOI for your dataset and keep it inactive so you can put the citation in a manuscript during the peer review process. After the article is published, just ask us and we’ll activate it.

Curation

A data librarian at the University of Iowa Libraries will also help curate your data when you deposit it in IRO. In addition to ensuring your file names and organization strategies are understandable to potential users, the librarian can also help you find open formats for your files, look at your documentation, and assist with the all-important metadata record that helps others find your work.

Get Started

Ultimately, depositing your data and code in IRO is a win-win. It helps you preserve your research outputs, disseminate them to increase the influence of your work, and enable scholars the world over to find and use your data and code for their own projects. And now that The National Institutes of Health require that researchers identify appropriate repositories for their data in their Data Management and Sharing Plans, IRO could be the answer – especially if no discipline-specific repositories exist in your field. IRO’s preservation, access, and curation features put it a step above other generalist repositories.

If you want help depositing your data or code in IRO or have questions about choosing a repository, Research Data Services is here to assist you. We have a guide on our website walking you through the steps to upload your content in IRO, do one-on-one consultations, and are available by email, too.

Dec 19 2022

Data Sharing: What are the Benefits?

Posted on December 19, 2022December 30, 2022 by Nancy Henke

Beginning in January 2023 the National Institutes of Health (NIH) will require data management and sharing plans for all awards that generate scientific data. Research Data Services at the University of Iowa Libraries has been supporting university researchers as they prepare for this change, with guidance on our website, informational workshops, and individual consultations.

With this new requirement, the NIH shows its strong support of data sharing, though the inevitable questions of “why?” and “what’s the point?” have been top of mind for many. So, what are the benefits of sharing data?

Research shows that data sharing has many advantages – for researchers, for the patients whose data is being shared, and for society more generally.

For researchers: A 2020 study published in PLOS ONE shows a significant advantage in citation numbers when data is shared in open access repositories. The study authors reviewed well over 500,000 articles published by PLOS and BMC and found up to 25.36% increase in citation counts when the associated data was fully open access – i.e., in a repository without restrictions on its use [1]. And sharing data also offers exciting opportunities for new research studies. For instance, analyzing old data with contemporary methods has led to novel “remixes,” such as a study that used weather reports in 19th century naval logbooks to model climate change, or one that analyzed text comments from patients in online forums to help clinical discovery [2].
For patients: Many have argued for the ethical obligation researchers have to make full use of clinical trial data, given that patients have donated their time and information in the pursuit of treatments and therapies [3]. Sharing clinical data widely could lead to the design of new trials, development of predictive models, and the creation of simulation tools [3]. Cancer researchers, specifically, have argued that because precision oncology is so heterogenous, no single research center can possibly produce enough data to adequately use models for prognosis and prediction. Data simply must be shared so there is enough of it to train the models and ultimately enhance patient outcomes [4].
For society at large: When those in developed nations with robust research infrastructure share their data openly, it allows those with limited resources – especially those in low-income countries – to use the data to investigate questions that are relevant to their own communities’ needs [2] [5]. Plus, there are major economic benefits to open data. A 2018 study by the European Commission found an estimated annual cost of €10.2 billion to the European economy when data is not shared [6]. The costs derive from inefficiency (especially wasted time), paying for extra licenses to access data, and storage costs.

Care must certainly be taken to minimize the risks of reidentification and other factors when sharing human subject data. Deidentification and sharing data through restricted access repositories are some of the ways in which this can be addressed.

In fact, participants in clinical trials are strong advocates for data sharing. A 2018 survey of participants indicated that they perceive the benefits of data sharing to outweigh the downfalls, and they noted that the most important advantage of data sharing was “making sure people’s participation in clinical trials leads to the most benefit possible” [7].

This belief in the value of data sharing actually extends beyond clinical trial participants. Pew Research found most Americans (57%) say they would trust research findings more if researchers made their data publicly available [8].

Ultimately, many stakeholders can expect to derive real benefits from the new NIH policy and increased data sharing, and researchers’ data management and sharing plans can lead to valuable contributions to the advancement of science and health outcomes.

If you’re tasked with writing your own data management and sharing plan, identifying a repository for sharing your data, or curating a dataset, Research Data Services can help. We have guidance on our website, offer workshops and trainings, do one-on-one consultations, and can answer your questions by email, as well.

References:

[1] Colavizza, G., Hrynaszkiewicz, I., Staden, I., Whitaker, K., and McGillivray, B. (2020). “The citation advantage of linking publications to research data,” PLOS ONE, vol. 15, no. 4, 2020 pp. 1–18, doi: 10.1371/journal.pone.0230416.

[2] Voytek, B. (2016). “The virtuous cycle of a data ecosystem.” PLOS Computational Biology, vol. 12, no. 8, doi: 10.1371%2Fjournal.pcbi.1005037

[3] Konkol, M., Nüst, D., and Goulier, L. (2020). “Publishing computational research – a review of infrastructures for reproducible and transparent scholarly communication,” Research Integrity and Peer Review, vol. 5, no. 1, p. 10, doi: 10.1186/s41073-020-00095-y.

[4] Vesteghem, C. et al. (2020). “Implementing the FAIR Data Principles in precision oncology: Review of supporting initiatives,” Briefings in Bioinformatics, vol. 21, no. 3, pp. 936–945, doi: 10.1093/bib/bbz044.

[5] American Psychological Association (2015). “Data Sharing: Principles and Considerations for Policy Development.” https://www.apa.org/science/leadership/bsa/data-sharing-report

[6] PwC EU Services, Cost of not having FAIR research data. Cost-Benefit analysis for FAIR research data. 2018. doi: 10.2777/02999.

[7] Mello, M. M., Lieou, V., and Goodman S.N., (2018). “Clinical trial participants’ views of the risks and benefits of data sharing,” New England Journal of Medicine, vol. 378, no. 23, pp. 2202–2211, doi: 10.1056/NEJMsa1713258. 

[8] Funk, C., Hefferon, M., Kennedy, B., and Johnson, C. (2019) “Americans say open access to data and independent review inspire more trust in research findings,” Pew Research Center, https://www.pewresearch.org/science/2019/08/02/americans-say-open-access-to-data-and-independent-review-inspire-more-trust-in-research-findings/

The University of Iowa Libraries

Library News

Author: nahenke

The Power of Persistent Identifiers in Data Sharing

Data Curation: Adding Value to Your Dataset

Why Iowa Research Online is an Ideal Place for Your Data

Data Sharing: What are the Benefits?