Where’s that data again?
The context: rich data
Large academic libraries, like ours at the University of Arizona, have conducted UX and assessment work for decades. We ask students and faculty for their feedback, ideas, and opinions through user interviews, surveys, focus groups, and other methods. We also gather behavioral data through methods like observations, usability testing, and diary studies. We have a rich history of gathering quantitative and qualitative data to inform decision making when it comes to everything from library hours to website improvements to redesigned services and spaces.
In libraries with mature UX practices, we’ve accumulated dozens if not hundreds of reports, slide decks, and spreadsheets containing findings from our research. A goldmine of data to inform user-centered decision making.
The problem: a disorganized mess
While librarians pride themselves on being organized, too often our research findings are stored inconsistently, making them hard to find or make sense of. At the University of Arizona Libraries, research is conducted across the organization, with departments having various norms and practices when it comes to documentation and file sharing. Some documentation is in shared file spaces such as Box, some is in Google Drive, some is on an intranet, and some is only on the researcher’s local computer. Some findings are in text documents, some in spreadsheets, and some in more feature-rich tools like Airtable and Notion. Some never made it to a digital format and exist only in the researcher’s’ minds or on sticky notes somewhere.
We often have data that can inform decisions, but don’t know it exists or how to access it. So we instead make decisions based on assumptions (no good), or replicate the study, duplicating effort in order to find out the same information we already learned but just forgot about.
The solution: a research repository
Fortunately, there’s a solution to this problem: a research repository. Research repositories are shared databases of research insights, and are increasingly common in large organizations with big research programs. I learned about research repositories in the Nielsen Norman Group course DesignOps: Scaling UX Design and User Research, taught by Rachel Krause. With more investigation, I discovered a whole community of people doing ResearchOps work to scale their design research practices.
So last year, our UX and assessment team embarked on a project to build our own research repository. Our goals of the repository are to:
- Store and archive research findings in one place.
- Make research findings accessible and discoverable to library employees.
- Empower library employees to use project findings from the research repository to make data-informed decisions.
The scope: keeping it manageable
In libraries and in academia, the term “research” is a broad one. Yes, we do lightweight user research and assessment all the time, primarily to evaluate and improve upon our services, spaces, and websites. We also do intensive, in-depth research that goes through IRB and aims to generalize findings to a broader population. Our librarians are faculty, so they conduct this type of “capital R” research to produce new knowledge and scholarship. We also collect data through systems and automated tools, like our library management system (Alma/Primo), web analytics (Google Analytics, Siteimprove), and occupancy counters (SenSource).
We need to make sure our research repository is scoped correctly so that it’s useful, practical, and sustainable. So we’ve defined our scope as:
- Data collected by the library that are of interest beyond a single department.
- User-focused research that aims to improve service quality. This includes projects that study user behavior, perceptions, and attitudes as well as people-generated data (e.g. interviews, surveys).
Outside of our scope are:
- Systems-generated data (e.g. circulation numbers)
- Ongoing data collection efforts (e.g. head counts)
The product: building a framework
We decided to use Notion since it allows us to create filterable, sortable tables of information. There was also a great article and template for using Notion to build a research repository. And Notion is a tool we were familiar with, as we used it already for note-taking and task-tracking.
Adapting the template, we began to input recent research projects. We made a page for each research project containing a summary of findings. We made the repository itself available through a public link (so that anyone could access it), but any personalized or identifiable information would remain behind a login (e.g. Box). See the public version of our research repository.
To adjust our template and refine our approach, it was helpful to add several research projects of different sizes and complexities. For example, we added results from a 5-minute survey while also adding results from a semester-long internship project that had multiple stages and contained a variety of research methods.
Ultimately, we landed on the following metadata for each research project:
- Title of project
- Start date
- Status (planning, execution, analysis, reporting, completed)
- Related department(s)
- Tags (for type of research, e.g. user interviews, usability testing)
- Products (e.g. main website, Weaver Library)
- User type
We also created a template for each project summary that includes:
- What we did and why
- What we found
- What we did next
- Now what?
Each project also has a sidebar of links for detailed documentation and raw data (e.g. Box) as well as related research that might also exist in the repository. We also have a spot to link out to associated recipes from The UX Cookbook, another project we’re working on to scale our UX practice.
What’s next: making it robust, useful, and sustainable
We have 14 studies captured to date. We’re slowly but surely adding all the findings from past studies that are still relevant and a useful reference. We have a backlog where we’re prioritizing research studies and assigning them for people to start drafting. At bi-weekly meetings, we’re reviewing drafts before finalizing them and checking them off the list. Notion’s ability to comment has proven helpful for the revision process, so we’re able to do much of our work asynchronously.
While early days yet, the repository is already proving useful. For example:
- An architecture student working on his senior capstone project reached out to us because he’s focusing on library study spaces. We shared two past research projects with him: cognitive mapping and a furniture assessment study. It was nice being able to just share a public link with him so he could access the summary of findings.
- We’re embarking on a design refresh of our family of library websites. Past research around information architecture and top tasks are proving to be especially useful reference points.
One of our next priorities is figuring out efficient and sustainable workflows. What is our content strategy? Who keeps things up to date? How do we work with researchers across the library to input their data? And how do we best communicate with staff when projects are added that might be useful for their work?
We’re also exploring a database of searchable “research nuggets,” where we add individual insights we think will be useful for future selves. We’re talking through workflows, scope, and taxonomy for this component of the repository.
We presented the research repository to library staff in January, 2021. We’d like it to be a library-wide effort, and asked staff to participate by using the repository to inform their work, submitting their research findings, and asking us to prioritize past research projects that will be useful to add.
It’s early days yet, but I’m confident our research repository will save us time, help our future selves, and lead us to more data-informed decision making throughout the organization.