Jennifer Abel
Research Data Management Librarian
University of Calgary Libraries and Cultural Resources
https://orcid.org/0000-0002-2294-7495
Jane Fry
Data Services Librarian
Carleton University
https://orcid.org/0000-0003-2223-1347
Nick Rochlin
Data Science Librarian
University of Victoria Libraries
https://orcid.org/0000-0002-0772-9342
Mathew Vis-Dunbar
Data Librarian
University of British Columbia, Okanagan
https://orcid.org/0000-0001-6541-9660
Introduction
The Digital Research Alliance of Canada (the Alliance), created in 2021, divides research data and digital research infrastructure support in Canada across three domains: research data management (RDM), advanced research computing (ARC), and research software (RS). A key challenge is building researchers’ skillsets across these domains, particularly in disciplines with limited or no formalized computational training. In May 2025, the National Training Expert Group (NTEG), part of the Alliance’s Network of Experts, piloted Jumpstart, a 20-hour national research data program aiming to tackle this challenge for early career researchers and graduate students. This free virtual program focused on connecting best practices in RDM with computational tools that enable best practices in scholarly integrity.
Why we did this
Jumpstart aimed to address two key elements in Canada’s RDM training landscape. First, RDM training in Canada is overwhelmingly characterized by ‘one-shot’ sessions, focused on Data Management Plans (DMPs) and data deposits. While valuable, this approach is disconnected from the role of RDM in the larger context of the research lifecycle. Consequently, it misses key concepts and activities that connect RDM to topics such as computational reproducibility and data integrity throughout a project. Second, there is a notable gap in learning resources for researchers with limited or no exposure to computational training in their disciplines, including basic concepts in machine- and human-readable digital content and the ability to work in non-graphical user interface environments. These skills are key to addressing issues of transparency in research and reproducibility of research outputs, adhering to the tenets of scholarly integrity, and, in many instances, enhancing workflows that benefit from automating routine tasks. While there are other, more computationally-focused workshops offered nationally (i.e., those offered by the ARC regional consortia), past feedback from participants indicates that those sessions can be overly advanced and too fast-paced for researchers new to these technologies. Jumpstart aimed to introduce participants to these computational tools, connect them to RDM, and, ultimately, integrate theory with practice at a foundational level.
Who Participated
The Team
The initiative was organized by NTEG, which at the time consisted of the four authors. To build and deliver the program, three graduate students from distinct disciplines were hired as research assistants (RAs) with funding from the Alliance. The team worked collaboratively, with the librarians handling the RDM-focused content (and the administrivia), the RAs handling the R-focused content, and everyone providing feedback. The RAs were crucial to Jumpstart’s success: their insight into the needs of graduate students and their ‘reality checks’ of the librarians’ assumptions allowed us to create a workable pilot.
Audience and Participant Selection
As we were somewhat uncertain about interest in the program, and concerned about the potential for low registration, the call for applications was open to anybody who was interested, with the note that graduate students would be prioritized in order to foster strong research practices in those new to research. The call was widely distributed across Canada, both within individual institutions and to the broader RDM community via national listservs. To maximize openness and inclusivity, Jumpstart was offered online and free of charge. Attendance was limited to 30 participants to create a personal experience and build a community-centred cohort in the online environment.
We received an overwhelming 175 applications within the three-week registration period. A full breakdown of the applicants, as well as other figures describing the applicants’ backgrounds and skills, can be found in Rochlin et al., 2025. Due to the number of graduate students that applied (88), and in line with our original intended audience, only student and post-doc applications were selected for a detailed review.
Applicants were asked to submit a brief letter of intent detailing why they were interested in attending the program, along with confirmation of support from a supervisor, and to indicate their familiarity with five RDM- and coding-related topics. Applicant assessment involved review of the letter of intent and the self-assessed skills, with lower skills being regarded more favourably during the review process to prioritize those with little to no RDM or computational experience. The selected cohort consisted of one 4th year honours student, 10 Masters students, 17 PhD students, and two post-doctoral fellows, spanning a wide variety of disciplines.
What we did
Jumpstart attempted to balance didactic introductions to the underlying theory and concepts of RDM with hands-on practical skills building. The latter was done using Open Science Framework (OSF) and the R Programming Language. The program ran over four-and-a-half days, with approximately four hours of content each day. A full overview of the content (available under a CC BY-NC license) can be found on the Jumpstart site.
The first day was largely lecture-based, introducing participants to the program, RDM, OSF, and DMPs. The second day focused on connecting RDM and DMPs to issues in reproducibility, including computational reproducibility. We then took a deep dive into R, which was meant to carry us through the subsequent two-and-a-half days, and included activities such as creating relevant documentation (readmes and data dictionaries), backing up content with version control using OSF, and tracing the provenance of data from the initial load into R through cleaning, organizing, and simple analysis. We also included a session on troubleshooting coding errors. The program wrapped up with a significant feedback session with facilitated breakout rooms that ran independent of those of us involved in the development of the program.
While Jumpstart was designed to minimize the participant learning curve, especially with R, we realized after the first introduction to R that our audience was clearly divided into three segments: students who found the training to be moving too quickly; those for whom pacing and content was just perfect; and the remainder who were looking for more advanced training. To address this, starting on the third day we broke the program into three streams and had participants self-select into a stream. Two of these groups followed the original planned content but did so at different paces, with the consequence that the slower-paced section did not cover as much material as its counterpart. The third group explored additional tools in R that help to support computational reproducibility: for example, file and package management, and automating RDM-related tasks through scripts. This pivot was not without its challenges and would not have been possible without the additional volunteer support we received from colleagues from across the country who had originally joined to help us with technical troubleshooting and support.
How it went
Based on our previous experiences, no-cost online programs tend to have 30-50% of the people who register actually showing up; for multi-day programs, attendance tends to drop off further as the program goes on. Thus, a key metric for success was participant attrition. For Jumpstart, 96% of the people (25/26) who were expected to attend on the first day showed up, and attendance averaged 85% throughout the week.
Participant feedback was solicited through both the facilitated breakout sessions on the last day and a short survey sent just after program activities finished. Overall, the feedback was very positive: in the 16 survey responses received, 88% of the respondents (14/16) said they were extremely or somewhat satisfied with Jumpstart, 94% (15/16) said they were extremely or somewhat likely to connect the program to their research activities, and 81% (13/16) said they were extremely or somewhat likely to recommend Jumpstart if it were offered again. In the breakout sessions and the qualitative survey questions, participants mentioned that they particularly appreciated
- the opportunity to learn skills that would help them in their research;
- the virtual and cost-free nature of Jumpstart, which lowered the barriers for participation;
- and the support and accommodations they received from the instructional team.
Areas suggested for future development included
- better integration of the RDM and computational/coding aspects of the course;
- finding the right pacing in terms of the amount of material covered, the time needed to cover it, and the varying skill levels of the participants;
- and creating a specific program for those working with qualitative data.
The program benefits were not only felt by the participants, but also by the RAs. In addition to gaining RDM, coding, and instructional skills, which they will be able to use in both their programs and their future careers, the RAs enjoyed the collaborative and welcoming team environment, which was quite different from their individually-focused graduate studies.
What comes next?
Given the number of applications we received, there is a clear need for Jumpstart to continue. NTEG is now working on addressing lessons learned and planning future directions (e.g., intermediate and advanced offerings, discipline- and audience-based offerings, varied methods of delivery), with the intention of delivering Jumpstart regularly in the future. The greatest challenge is securing funding to continue delivering the program. While a significant portion of Jumpstart was developed with in-kind contributions (e.g., the authors’ salaries from their institutions; additional professional volunteers during the program; infrastructure for delivery, such as Zoom licenses), RA engagement was a critical component requiring compensation. NTEG is continuing to work with the Alliance to ensure that Jumpstart is properly resourced in the future to allow Canadian researchers to gain the crucial skills they need in the modern research landscape.
Acknowledgments
We would like to acknowledge the funding provided by the Alliance to hire our three fantastic RAs: Maria Amoros (University of Victoria); Gia-Huy Hoang (University of Calgary); and Erin McCoy (Carleton University). As well, thanks to our volunteers who helped in many other ways during the week to make the pilot Jumpstart a success.
References
Rochlin, N., Abel, J., Fry, J., & Vis-Dunbar, M. (2025, June 4). Laying the Foundation: A Pilot National Research Data Bootcamp. IASSIST 2025, Bristol, UK. Zenodo. https://doi.org/10.5281/zenodo.15684148.
