Jordan Kaufman
Assistant Professor and Research Data Librarian
University of Tennessee, Knoxville
orcid.org/0000-0003-3540-1421
Introduction
According to a 2022 report from the Data Literacy Project, “data literacy is essential to the future of work”. One practical component of data literacy in higher education is research data management (RDM). Yet, RDM instruction across universities has often developed in decentralized ways, such as, shaped by individual colleges, courses, and librarians, resulting in a mix of discipline-specific approaches and generalized best practices (Xu et. al.(2022).
Introductory RDM sessions can rely heavily on slides, terminology, and technical language. For newcomers, concepts such as metadata, documentation, and file structure can feel abstract and disconnected from their daily work. When foundational instruction feels overly technical, it risks reinforcing the idea that data management is complex, specialized, or only relevant at later stages of research.
Instruction Design and Why
As the new Research Data Librarian at the University of Tennessee Libraries, I considered how to build accessible entry points into our growing research data services. Like other library service areas, RDM support can take multiple forms, such as LibGuides, one-on-one consultations, and workshops. I created an interactive workshop designed specifically for beginners that would make abstract concepts tangible and reduce anxiety through collaboration and playfulness.
1st Iteration Workshop: Data Management 101
The 1st iteration of the workshop was constructed for an internal library’s research interest group. The workshop began with a brief framing of data management and why it matters. From there we moved into the interactive activity, and closed with reflections and takeaways. The activity started with an introduction to basic spreadsheet structure. I brought my dog, Grimm, to serve as a “live” data point. Using a whiteboard, we built a spreadsheet together in real time. Grimm represented a single observation (row), and participants worked collaboratively to determine what variables (columns) we should include to describe him.
Some variables were straightforward: breed, hair color, age, and weight. As we added each column header to the board, we discussed what kind of data it would contain (text, numeric) and how values should be recorded for consistency. (For example, should age be recorded in months or years?)
I then pushed the group to think beyond the straightforward observations. I promoted questions such as: should we include Grimm’s favorite treat or his preferred toy? Participants debated whether these were meaningful variables, how we would define them, and whether responses should be open text or limited categories. Those decisions naturally surfaced larger data management concepts: scope, standardization, and anticipated future use.
By collaboratively constructing the spreadsheet, participants experienced how quickly ambiguity and inconsistency can emerge, even when describing something as familiar as a dog. The activity created a shared reference point that we later used to introduce documentation and metadata.
2nd Iteration Workshop: Data Management Doesn’t Have to Be Grimm
The 2nd iteration of the workshop was constructed for Love Data Week.
As I reflected on the 1st iteration of the workshop and took in feedback I opted to add more variety to the spreadsheet activity and expand the metadata section. For the activity, I brought in a variety of physical balls for Grimm to choose his favorite toy. This gave the group the opportunity to dive deeper into how values are determined and described. For the metadata section, I introduced the distinction between variable-level documentation through a data dictionary and project-level documentation through a README. The expansion of the spreadsheet activity allowed the group to develop and create a data dictionary together. I was able to demonstrate how ambiguity emerges without documentation and made metadata visible and structured.
By instructing multiple iterations of the workshop, I created clearer alignment between the learning objectives and the activity, such as reiterating why RDM matters, learning why documentation matters, and how to document.
Feedback and Future Directions
Initial feedback from participants highlighted the value of the workshop’s interactive and low-stakes environment. While designed for those in the beginning phase of their research, these early sessions functioned as pilot environments with library participants, creating space for open discussion and reflection on foundational concepts that are often assumed or overlooked. As one participant noted, “The environment cultivated in this session was amazing. It felt like we could be open and honest and ask real questions. Thank you so much for offering it!” This response highlighted that approachability and psychological safety are critical components of effective introductory instruction.
Feedback from these sessions also informed several refinements. In early iterations, documentation was introduced primarily through the concept of a README file. Based on participant engagement and areas of confusion, I expanded this section to include a data dictionary, providing a more structured approach to variable-level documentation. I also incorporated more concrete examples, such as defining a variable based on Grimm’s toy preferences, to help bridge the gap between abstract concepts and practical application.
As I continue to develop and scale this workshop, future iterations will focus on adapting the content for broader campus audiences and exploring how the session can be embedded within disciplinary contexts.
Conclusion
Research data management becomes less intimidating when learners are invited to learn about it in a low-stakes environment with simple concrete examples. Rather than presenting best practices as a checklist to follow, my goal was to make foundational concepts clear and manageable, while helping learners recognize the complexity involved in doing them well. As participants deliberated how to record Grimm’s age or define his preferred toy, they demonstrated the same challenges researchers face in more complex datasets: ambiguity, inconsistency, and the need for shared understanding. The playfulness of the example lowers the stakes, but it does not diminish rigor. On the contrary, it creates space to focus on foundational documentation principles, such as: clear variable definitions, constrained values, and transparent decision-making, without the pressure of disciplinary expertise.
What begins with a dog and a whiteboard ultimately leads to a deeper understanding of how thoughtful documentation makes research data usable, shareable, and sustainable.
