The following is excerpted from an invited talk I gave at the Oklahoma State University Library in early April. I’ve embedded the slides below but you can also scroll through for my notes – not the most thorough writeup I’ve ever written but enough to give you a glimpse at my discussion points!
I’d love to hear whether these ideas fit with your experiences with library digital scholarship services, either here in the comments or on Twitter.
Today I’ll be sharing the secrets I’ve picked up on during my work with data services. They’re not secrets in the sense that there’s anything remarkable or unusual about them, rather that this is just such a new area that there’s little information about how to get started with research data management. It can often feel intimidating and overwhelming.
But the reality is that we’re all generating more digital stuff, more data, than ever before. Collectively – universities, departments, labs, individuals – have no clue how to manage it but that won’t stall the pace that we’re generating data.
Libraries, with our long history of information organization and preservation expertise, can be great leaders in this area. Here are a few ways I think we can assert value in this space.
We can share our knowledge of various research data management concepts with researchers at every stage, help them understand how to be more open in their research workflows and publication, and explain how to comply with federal funding requirements like the Office of Science and Technology Policy’s memo stating that publications and data resulting from many agencies’ funding must be made openly available.
Going further, we can build concrete services that move beyond just providing guidance. We can develop teaching and training programs targeted toward our keenest institutional objectives, whether that’s data information literacy training for graduate students or librarians embedding in labs for more in-depth projects. Libraries can have a role providing infrastructure to enable research – storage, collaboration, and preservation – and then partner with researchers for the resource-intensive act of curating data that has been collected. Libraries don’t need to fill every one of these roles, particularly when it comes to technical infrastructure. Often these can be campus partnerships; however, libraries can vie for a place at the table in order to to shape how we approaches the multitude of options for investing in new services.
And then there’s the ever-important policy piece. It’s a little less flashy but certainly critical to have librarians engaged in influencing policy related to research data management at every level.
Consultations! Infrastructure! Policy! There are a lot of possibilities but somehow it seems like we’re still in a whole new world, a vastly different terrain that what we’re accustomed to. It’s a little unsettling to be sure.
Before we get too deep into talking about RDM, I want to share a bit about the context I’m coming from.
I work at the University of Wisconsin. I like to say that I inherited Research Data Services – it had existed since 2011, started as a grassroots effort by a passionate librarian who wanted to help equip researchers with ways to meet data management plan requirements. However, by the time I arrived in 2014 the path forward was uncertain.
RDS is comprised of knowledgeable people from across campus – libraries, IT, research groups – but the crucial part to note is that everyone volunteers their time on top of already busy jobs. We help with consultations, DMPs, and training and education. We existed: we had a name and a website and we met every month; however, the vision for what this group was supposed to do beyond waiting for a consultation request to come in was murky at best. There were plenty of times when I thought to myself, “I have no idea how to make this work.”
I had some starts and stops. I had to acclimate myself to a slower pace of change and test out my own capacity as a leader. But pretty soon, I stopped waiting for a sign of approval and decided to just try things. I knew if I veered too far beyond what was wanted I would be reeled in. I claimed responsibility for RDS. I took on a more authoritative role (while still recognizing the delicacy of my authority – again, I have no direct reports). I directed projects and built relationships and gave my best energy and kept my own spirits up. And now I’ve been at UW for two years! It’s interesting to reflect back on what we’ve been able to try during that time.
Connections and community
It took me some time to understand various organizational structures at UW. Figuring out how the campus is organized is hard enough – and then there are our fifty libraries. I learned about units, individuals, rivalries, partnerships, deep histories. I fanned the flames in the library, growing awareness about research data management and the broader impacts of digital scholarship. I went beyond my library and met with IT and research support staff across campus. I talked to researchers, began picking up on their practices, needs, and pain points. I engaged RDS’ sister groups on campus, the Digital Humanities Research Network and the Open Meetup. This relationship- and community-building work is no easy feat. Sometimes I’ve felt awed by the amount I’ve given myself over emotionally to this work but the worthy reward is feeling so much more connected to this university.
A stronger foundation for RDS
Alongside building connections, I set out to lay a stronger foundation for Research Data Services. A practicum student who I subsequently hired on, Cameron Cook, was a huge boost to what we were able to accomplish. We improved our file organization and sharing practices; we created and improved project documentation. We renewed processes for tracking and assessing consultations. We refocused on design, considering both print and digital, including a sizable website redesign. We gave new energy to our content creation workflows, revamping the blog and starting a monthly digest to help mobilize our content. We remained active on social media, particularly Twitter.
So now I am increasingly connected to this campus and we have a solid, cohesive vision for how we are promoting RDS services. The next step I’m envisioning is moving beyond just providing guidance to developing greater services – even if that just amounts to building capacity for teaching and training by looping in more interested librarians under a train-the-trainer model.
Teaching and training is at the top of my list, including “inreach” to librarians, expanding my workshops for grad students and early career researchers, and taking a greater role in organizing Data Carpentry workshops. I want to weave research data management into broader discussions on open science and reproducibility. I’m a Center for Open Science ambassador now with an eager partner ready to start bringing more of a spotlight to these ideas on campus.
Data services is not a one-person role. There’s widespread agreement on that one, I think, but how you grow partnerships that move projects forward is a whole different animal. Implicit in every new direction I’ve just listed is an extension of existing campus partnerships that have been growing and developing these past two years: with my fellow librarians, with the RDS consultants, with the Advanced Computing Initiative, Office of the CIO, and so on.
Last, it’s important to note what I did and what I want to do, but it’s also exceptionally important to note what was off the table. Wearing many hats, being chronically underresourced, and getting pulled in many directions means you are making choices every day about where you will put your energy. You can only hope to continue focusing your energy in a meaningful direction. You’ll be disappointed, disillusioned, and feel like you are not doing enough if you fixate on what you are not able to do. I fell into this trap for a long time. One example of this is that I would get users asking me about where they could deposit their data. I had to tell them time and time again that our institutional repository was not a suitable place for that. I was constantly the bearer of bad news and I felt powerless to help them. When I started to recognize the toll this was taking on me, I knew I needed to set some things aside. I needed to start thinking in years rather than months, for one, and I needed to do what I could control. One bite at a time. This doesn’t mean I’m not advocating for more sweeping change and resource allocation, but I’m getting less bummed out about what I can and cannot control.
Now that you know a bit about the context that’s helping frame my perspective, let me share my recommendations for programs that are just getting started, or need a new burst of energy:
What I mean by this is finding doable ways to assert that the library is exploring this space all while you are concentrating resources on staff training and easing into research data management topics.
Part of this is clearing the path: making it easy for people to connect the library with RDM.
What do you want your organizational structure to be?
One consideration is whether you are keeping your project team(s) separate from your core consultant group – generally, I recommend this.
Develop a website that articulates your services, if you choose to offer them, and information for the RDM-curious. A major project I undertook this past year was an overhaul of the website design, though I was lucky because RDS had really exceptional content from the previous website. We just needed a refresher.
I’ve also found consistent branding to be a worthy investment. We use the free version of Canva to create our flyers, handouts, and social media graphics.
Once you’ve settled on how you’ll represent and promote your data services, the next step is to dig in.
Get more comfortable with data consultations. It can be tricky: you don’t know what type of question will be thrown at you, but you have to go in with the knowledge that you’ll be able to help find the answer. It’s really just a reference interview, just with a different focus.
At UW-Madison, RDS hosts a monthly speaker series every semester. This series is useful in a few ways: it creates relationships between researchers of different disciplines and the library, and allows library staff and beyond to explore RDM and emerging research practices in an informal environment. I learn a ton about disciplinary research practices at each talk. This is by far one of the undertakings with the heftiest payoffs, in my opinion.
I also had great success bolstering library staff engagement on RDM through a reading group (read more about that here).
All data librarians know that 70% of your job* is explaining and justifying your job to various audiences, at least when you’re a new unknown entity at your library. I think we need better strategies for talking to different groups about RDM. Elevator speeches can do that.
*highly unscientific stat
We worked on elevator speeches at the Midwest Data Library Symposium. I am adapting some of that content, including the section on elevator speeches, for an August research data management bootcamp at UW.
(See all MDLS material here.)
There’s only so much talking you can do about data! At some point you have to get your hands dirty.
The University of Minnesota Libraries did just that by initiating a library-wide data curation pilot. They gave us the gift of a helpful writeup too.
Design thinking is a way of reframing challenges. It has been a hot topic recently, for good reason – I think it has a lot to offer us as we consider how to tackle increasingly complex issues.
There’s even a design thinking for libraries toolkit, which I highly recommend checking out.
Ultimately, what’s most important is working together through all of this. We have to build capacity together. AND NOT JUST ONE TRAINING: a dedication training program over time. Libraries should incorporate research data management and scholarly communications expectations into job descriptions, including adapting existing jobs. This has to happen in conjunction with a lot of communication and training but it is a strong path forward. Frankly, having one data person is not scalable. My role straddles that of a strategist and functional specialist. Ideally, multiple functional specialists can work in tandem with subject librarians to create a strong network of data support. Subject librarians often have deep ties to their departments and research areas; they are critical partners in sharing research data management concepts. And in understanding how disciplinary differences affect research data management! Roles are shifting. They should shift. They must shift.
This is the moment when I give my pep talk: Don’t fear failure in any of this. I mean it. Whether nobody shows up for your RDM workshop, whether you misunderstand and accidentally misrepresent an RDM concept. It’s a messy space but we only learn by wading into it.
Grad students, in particular, need librarians willing to go there. They are often the keepers of data in the labs, plus as emerging researchers and professionals they need these skills and are ready to acquire them.
That’s why I started a workshop series for graduate students and early career researchers; one of the topics was RDM (view those slides here).
We have to make these topics fun, if we can. RDM is intimidating enough that we do everyone a favor by injecting a little lightheartedness. This doesn’t mean detracting from our professionalism, just making a very bland, dry, guilt-inducing topic more accessible. Think about how messy and disorganized the files on your computer are. You feel uncomfortable, right? This is exactly how our researchers feel when we are talking about RDM best practices: they’re imagining the hot mess that their own workflows produce and for many that means shutting down a bit so as not to feel judged. By coaxing in a little humor we become more relatable and we increase our odds of success in actually changing behavior – not just spouting off information that goes in one ear and out the other.
I lucked out with an assistant who is a skilled cartoonist and here’s one example of what we tried to inject a little fun: data comics.
So let’s be bold!
- Embrace discomfort and an uncertain path
- Keep sharing openly about our successes and failures
- Approach RDM within a broader context, connected to research workflows
- Work on our elevator speeches
- Stop letting commercial entities take the lead (prevent vendor lock-in!)
- Not fixating on perfect digital preservation over workable solutions
Change is bewildering, thrilling, petrifying, overwhelming – depending on who you are and how you feel about RDM. But there’s a place for everyone through these transitions. RDM is the great equalizer. Doing it well means developing skills that cross boundaries. Beyond allowing us to impart new skills to our university community, it’s an area that can also act as a sandbox for improving relationships between library colleagues, too.
We’re in this together.
The future of data services? We’re going to see teams continue to come together on local, regional, national, and international levels. We’ll see institutions develop data information literacy programs, either integrated with or alongside traditional info lit programs. Services will become aligned with computational services on campus – high throughput computing and programs like Data Carpentry. Disciplinary interest in open science and reproducibility will usher in new audiences for RDM training. We’re going to see a continued emphasis on how federal funding directs the conversation and services developed (for better or for worse). For institutions that embrace their roles in RDM and engagement in the research enterprise, we’ll see further support for research workflow platforms. And finally, investments in the curation and preservation of any worthy research data that survives (openable, readable, understandable).
If only building to that future were simpler, amirite? Ideally we have various levels working together – individuals, units, institutions – to build that future but in truth it often feels more disjointed than that. It’s a murky, murky world and that can lead to wanting to wait for more clarity. Pausing, paralyzed. Or feeling constantly like the efforts we make aren’t good enough. For individuals without the power to corral resources, I say the answer is often just to start something. Anything. I know in my experience, often my efforts do not feel like they are enough. That’s why I give these talks: in the hopes that it will shine a light on what else we can do collectively to support each other and to support this area appropriately.
It does take time, after all. I remind myself of this often but I still poke and prod to get things moving.
We make our lists.
Defining RDM success, is, I think, still highly dependent on institutional culture and objectives. But what I know for sure is that doing something, starting somewhere, is always more useful than waiting and not taking a direction on data services at all.
If you’ve made it all the way to the end of this post, I’m impressed. I hope this was helpful and I’d love to chat further if you have any questions or suggestions for me as I continue to grow data and digital scholarship services at my institution. You can view my complete slide deck for this talk here.