Recording and Notes from our September Live Update

29 Sep 2020

We had a wonderful audience for our webinar on September 22. At one point we had 135 attendees participating, and our Q&A session was full of excellent questions and thoughts.

A copy of the slides from our presentation includes notes explaining what we shared.

We also have a video of the Question & Answer session. See below for a full textual summary of Q&A content.

Attendees

At the beginning of the webinar, we asked attendees to complete a poll to share some details about their connection to visualization in libraries. We had over 100 responses to the poll. Responses below show that the majority of the audience works in libraries, though not necessarily in a visualization support position. The attendees were most excited about sharing techniques and best practices, though some would also like more visualization training.

What best describes your current work situation

Response	Pct	Bar chart
Academic, research or special library	77%	`+++++++++++++++-----`
Other	5%	`+-------------------`
Public Library	4%	`+-------------------`
Non-profit	4%	`+-------------------`
Education (primary, secondary, academic)	4%	`+-------------------`
Government	3%	`+-------------------`
Student	2%	`+-------------------`
Freelance visualization profession	1%	`+-------------------`
Industry	0%	`--------------------`

My primary role is to do visualization (i.e. public services) in a library

Response	Pct	Bar chart
No	78%	`++++++++++++++++----`
Yes	22%	`++++----------------`

What resources would be most helpful to you?

Response	Pct	Bar chart
Shared techniques, best practices	41%	`++++++++------------`
Training related to visualization	29%	`++++++--------------`
Community of practice	15%	`+++-----------------`
Shared instruction materials	13%	`+++-----------------`
Mentorship	2%	`+-------------------`

Question & Answer Session

We had a lively question and answer session. In addition to the video we recorded, we have also compiled information from the Q&A box, chat, and presenters in the summary below. The questions have been loosely grouped by topic.

About Visualization in Libraries

Did you have specific viz roles in the library or were these added on?
- Andy: There’s an interesting way that my position got created here at the University of Southern California. A few years ago when I first started at USC I was actually a, the full title was an Interdisciplinary GIS Library Fellow, which was a two-year contract position, and in that role i actually kind of sketched out GIS sort of needs. And one of the things that I also started talking about when I was sort of giving an assessment on that position was that the need here was much more, it was much more than just GIS. It was, there was a lot of work with just data and using a lot of other tools beyond GIS tools with students. And it took a little while an additional sort of year or two before this sort of visualization specialist role fully got developed and created, but it was specifically, you know, a visualization specialist working with with the USC academic community to do visualization support. And it’s really sort of broadly sort of developed and then I’m sort of allowed to sort of cater the role to specific groups on campus that I work with. So I really kind of like that.
- Justin: I can go next. So I… my role is, yes, I’m the visualization librarian so my focus is really on data visualization. When I started as a librarian in 2011, maybe, at the University of Michigan, I started also as a GIS librarian and my focus was on spatial data. And a couple of years into that, then, Angela got her position at Duke and Duke had posted for a visualization librarian and I was, you know, seeing more and more questions about data visualization. I was getting really excited about it, and so I was able to transition my role from GIS to visualization and actually as soon as I did that I arranged a little trip and I went and visited Angela and some other people in the Research Triangle to sort of learn about what they were doing and their work and maybe that was actually where we first met.
- Angela: Yeah, I think that’s it. I think, yeah, Justin came up for a visit. My position at Duke started in 2012 and was billed as Data Visualization Coordinator, which is essentially a public services position to build a data visualization services program at Duke and to focus on things like consultation, instruction, some infrastructure, some guest lecturing, things like that. And Duke saw a lot of good uptake in this area. It blended really nicely with our data and GIS department at the time, and we developed a a community that had enough interest in visualization support that we were able to argue for a second position in data vis support. So for a while there we had two visualization librarian-style positions, although it wasn’t quite that name, but certainly consultants. And then a couple years ago i actually transitioned to an assessment position in the library. So I’m still at Duke and I still do visualization work, but I am more focused on internal visualization needs and, in some aspects, reporting and data warehousing for for Duke libraries. So we kind of take advantage of visualization best practices in our own work. So, I think, yeah, the process of starting a position as a visualization librarian was very, it was very intentionally part of my position to focus on visualization but since I have transitioned to a position where visualization isn’t exactly the primary objective but sort of a dovetail with the primary objective of reporting and assessment, I think it’s given me a lot better perspective on what it’s like to have visualization kind of added to an existing position in a library, instead of be the primary focus.
I’m curious about how folks have integrated visualizations into library catalogs as a way of recontexutalizing library collections for students
- That’s an excellent question. These are the kinds of things that I think the Viz and Tell will be great for because frankly I have not seen catalogs that include a lot of visualization embedded within the catalog, so that would be something I’d push back to the community and see if anybody else has has seen that, I think that’d be really amazing. I think where I’ve seen visualizations as kind of an assist to information searching and browsing is maybe more in like the citation index realm where I feel like Web of Science or some of those tools have sometimes had embedded visualizations to look at the citation network of the thing that you’re looking at. But I don’t know of any library catalogs where there’s kind of embedded visualization other than like I guess sometimes I’ll see on a repository visualizations that summarize an author’s citation profile over the last few years or visualizations that show the Altmetrics data for a particular author. I know we’ve got embeds on our repository that do that, but I don’t know if we have any really good, like, subject browsing visualizations in library catalogs. Yeah, I mean, I think ripe for development here. Next IMLS grant, visualizations in library catalogs.
- An interesting question to start asking all the ILS vendors
  - I agree!
- Green Glass (OCLC), but behind the seen, HathiTrust
- Yeah, was thinking of altmetric donuts
At a small college library like ours, we don’t have a dedicated data/vis/assessment librarian. Are the modules, examples, and upcoming sessions going to be accessible/relevant to librarians for whom data visualization may seem peripheral to their main role as a librarian?
- Angela: And I also feel like this question kind of touches on some of the other concerns, like with public library usage and resources. So I think I might try to address a couple of these questions about, like, how can the work of this grant really be utilized by people in a lot of different situations. I hope that’s fair to do that compression. And I think that’s that’s where we feel very strongly about making all of our stuff public. So we have a bunch of things that we’ve shared from our August meeting, even so you can what our process was for deciding on what we’re spending time on and then in the development of these instructional modules, of course, the idea is to share as much as possible. I think at the very bare minimum our intention is to share the slide deck and kind of, maybe a speaker guide or a presenter guide. But I think our hope is maybe even to be able to record some of these modules so that they can just be shared with patron groups without the burden of someone having to take on the instruction role. So in these kind of asynchronous times, sometimes people will sort of “flip” the classroom and share a video that is the lecture and then host online discussions about that lecture material or something like that. So that might be an opportunity for smaller institutions or for public libraries to share the videos and the instructional materials directly and then organize some other kind of reflection on those materials.
- response combines this question with “I wonder if there’s any hope for public libraries…”
- see also “There are quite a range of tools…”
- Thanks, Angela. That’s all helpful.
I wonder if there’s any hope for public libraries, given the limits on resources. Covid has moved usage from buildings to online, but assessment and reporting is traditionally focused on gate count and similar.
- See answer to “At a small college library…”
- I’d really love to think more about this and see if there are ways to help public libraries and other libraries with fewer resources, especially for assessment and reporting. Feel free to stay in touch if you’d like to talk more about this, or join in the November discussion!
This project is very exciting, kudos to everyone! I was wondering about how this project can connect to the teaching of viz in library science master programs, which is in many cases a bit lacking.
- Justin: I think it’s a pretty complicated question, especially given how quickly the MLIS landscape and libraries are shifting. But I can just share a little bit about what’s going on at U of M, in case it’s of any help, and maybe some of the the other panelists can share what’s going on at their universities. But the University of Michigan, our Information School used to be predominantly a library school, but it is very rapidly you know, sort of, starting at the time when I was there and then over the past decade, shifted into being really like an information school with a focus on HCI and data, you know, and data science and these kind of things. And so there’s still a library science degree that’s offered there but a lot of the visualization stuff is happening in the data science side of things. And so it’s really difficult, I think, to kind of bridge that gap. So I do a couple sessions in the library science program, sort of talking about visualization, and then I’ve also over the past years, although it’s kind of shut down right now due to the pandemic, had an internship program where I had usually five to six students who would work on data visualization projects with me and I’d always get a nice sort of mix of, like, half the students would be, kind of, library science type students and then the other half would be students on the more data science side of things, oftentimes students who are more interested in working in non-profits or kind of going a little bit beyond the kind of standard industry work. But yeah, I think it’s a challenging thing because of the, sort of, the economics and you know what information schools are trying to do but I think that there’s some really important work to be done in sort of integrating this work into library science programs.
- Agreed, I would love to see more targeted courses on data services in library schools. Maybe data vis-specific courses are a stretch, but broader data services courses might work.
- Thanks for this! I’m a prospective MLIS student. But given Kahlila’s point, is it more prudent to get a degree in data science and then apply to data-related librarianship positions?
- Justin: You know, I’m sure I don’t have the right answer for that, but I think one thing that works for me and I’ve seen has been very effective for other students that I’ve mentored has been to do a library science degree but then take as many cognate courses in data science programs. Or you know, like, I took a GIS, a couple GIS courses, and that was really a sort of a launching point for me. So, you know, I think it’s good to… it’s good to be integrated into what’s going on in libraries, so I wouldn’t, you know… if you want to work in libraries, I wouldn’t abandon the MLIS, but I do think you need to sort of expand your education, you know, into the more sort of data and design type courses that sometimes aren’t offered in MLIS programs.
- Andy: Yeah, the only thing i would add to that, too… And it is, sort of… I know I’ve seen more and more positions that kind of walk away from that library degree requirement when there’s sort of, especially when there’s more of a robust team or already infrastructure in place at a library. So when there’s a larger sort of team in place and, you know, there’s a data or GIS position, I think libraries are much more adaptable now to hiring people with those skills because they know I think the reality is that it’s really hard to come by those skills. So, it really depends on the institution. And then I know that, you know, there are a lot of people on this call that are at smaller institutions where, like Justin was saying, I think that there’s a real balance where you still need to really have that library knowledge and that library degree is still very very valuable for those institutions. And then just having the experience of classes or a certificate or some other sort of way to kind of work on that, developing those skills, is enough. So it’s hard, actually, I think. It really depends where you’re looking.
- Justin: And that reminded me of just one more small thing, I mean, I think that because this sort of work is so emergent right now, I guess, to use a sort of a catchphrase, that really internships are, like, incredibly important. iI think that when I see people sort of move into this kind of work, the launching pad is… it really tends to be, kind of, like, good hands-on experience working with other people who are doing this work.
PR and assessment share common goals but they have different goals. What are most common goals of visualization in your project?
- I agree, assessment can often be used to promote services or advocate for more resources. Visualizations in assessment projects are often designed for either decision making (get a quick glance and see if the performance is as expected) or for outreach (communicating successes and obstacles to stakeholders).

About Teaching Visualization

There are quite a range of tools to create visualizations. One challenge in teaching is determining if we need to teach specific tools, which tools are most appropriate for the discipline, which tools are the best to support.
- Justin: When we initially started that was this sort of like… “teaching beyond the tool,” I think was a phrase maybe that Angela came up with that we used a lot. But you know one of the things we really sort of saw was that a lot of the kind of, the initial work around visualization in libraries was very very tool specific. It was heavily tied into, you know, using maybe R or Tableau or even Excel or something like that, and we really wanted to, especially, you know, as the, sort of, the tools and technologies start changing so quickly and also, you know, some places, some of the larger libraries, you know, you have people who can be dedicated to this and there are also… and then, you know, like someone was saying, at smaller libraries it’s sometimes added on to a social science data librarian position or something like that. So really we tried to think about sort of extracting, kind of, teaching principles, teaching, you know, through examples, teaching design, and these kind of things in a way that hopefully, both as instructors and then for students, they can take that knowledge and sort of migrate from one tool to another and that it can sort of provide a framework for either teaching or further learning. You know, you can learn some principle and you can figure out how to do it in R but you can also figure out how to do it in Stata. So that question that you have, Stephanie, is I think really at the heart of what we wanted to do with this grant and the, sort of, the knowledge we wanted to share and sort of support.
- Angela: I think, yeah, I mean, honestly my approach to teaching beyond the tool, I really credit Justin with because it was one of the first things that I think Justin and I discussed was, like, it’s very easy to teach a tool… to “teach a tool,” but it’s kind of incumbent on us to think past that and to think what assumptions the tool is making, what values it’s bringing into the conversation, and how we interact with the tool. And so I think this blended approach of teaching what the tool can do but also teaching what is possible in visualization outside of the tool and how to sort of negotiate with the tool to produce the visualization that you really want or that’s really appropriate. I think incorporating non-tool-based instruction into every tool-based instruction session is the way to really get people more comfortable with these kinds of visualization concepts.
- Andy: I think maybe the only thing I would add to that, too, in terms of thinking about tools… and I think initially when we were talking about all this, too, is how much the, sort of, the tool landscape now has become so complicated with, sort of, especially in GIS world with, sort of, more corporate partners. And just in the amount of time we’ve been working on this grant, we’ve seen a lot of tools that we’ve used go behind paywalls or go behind subscriptions where all of a sudden, where before… I’m going to be teaching Carto this week to students, for example. Before, the sign-up process was very easy. Now there’s you know, this partnership with GitHub and you have to go through those steps and kind of do that. And so again kind of thinking about stepping outside of just the tools and kind of getting more into concepts, getting more into sort of workflows, documentation, thinking about data more, even, just seems like… just a way to sort of, you know, kind of step back and frame our work a little bit more and put it all into context and kind of… and I think also a lot of students are just really really interested in sort of the ramifications of choosing you know, an open source tool, for example, over a proprietary one and all the implications of that kind of of work. So that’s the other, I think, strand in that tool story.
- It might also be helpful to discus how we as a community plan on keeping these tool tutorials up to date, when there are updates to software which can change how we need to teach or supportt tool use.
- This reminds me of supporting citation management tools and needing to select a few to support and then continually stay up to date for teaching purposes.
- Agreed. Especaily with tools like Adobe or ESRI, which are now moving to be automatic updates online.
What theories or theoretical frameworks do you each use to situate your reseach?
- Angela: I will say my position does not include a lot of research, and when it does it’s less focused on theoretical frameworks than on kind of… I won’t say pure science, but I’ll say kind of exploring behaviors and opinions around information visualization. So I would say I take a more methodological approach of, kind of, usability studies or learning about perception with visualization through quasi- psychological experiments. I think Justin and Andy are more on the theoretical side.
- Justin: I can say for me, I… a few years back as I was working on this I discovered the work of Jacques Bertin, I’ll put it in the chat… he was a French cartographer and in the ’60s he wrote a book called Semiology of Graphics, which is this amazingly weird book where he tries to sort of break down kind of like everything that you can show in a visualization. And I love, even though it’s quite old, you know, I love teaching it because it really sort of gets students thinking about all the different choices they can make. And, you know, and there, I mean, there are lots of things out there like Stephanie Evergreen’s work, there’s, you know, Tufte… and I think because so much of my work in visualization is around teaching I just sort of pull bits and pieces that really speak to me that i find are easy to explain to students, but one of the things I sort of keep coming back to and the things that I really like about these frameworks when they’re successful is that they get students to think about the choices they’re making. And I think when things are really driven by the technology, it sort of makes choices for you. So just trying to get students to think about, you know, why’d they make something green, why’d they decide to use size versus color to communicate something. And I think that that really… I guess not so much research, but teaching.
- Andy: Yeah, I mean, I can quickly add, I feel like, I… for me personally, lately, I’ve really I’ve really been looking towards Data Feminism by Catherine D’Ignazio and Lauren Klein as a book. And I feel like there’s so many, I mean, just really nice clear pieces out of that work that we can kind of apply and sort of resonates… I’ve been recently working a lot more with undergraduate students and just sort of talking about labor in data work and visualization work and things like that were kind of… I don’t know, like, really making some of this just more tangible and really, I don’t know, kind of situating it more in the real world sometimes, where you kind of have not so much abstract kind of visualizations out there but they’re kind of, you know… you see something in the New York Times or you see something… and it’s sort of not connected to the people that, you know, a visualization might represent in the real world or who created it. And I think work like Data Feminism kind of helps to to put things into context. And I think that’s where for me right now is kind of really finding readings finding sort of theories that that can kind of really make the work tangible and we can kind of communicate it and really make it understandable to students and engaging
Have you surveyed or assessed skill/competency levels versus self-estimates of skill for data visualization? I’d love to hear about that.
- Not as actively myself, but I’m very interested in this! We’re trying to think more directly about assessment with each of our modules.

About Grant/Project Work

Would the Nov session include library visual assessments from collection analysis? Or is this focused on instruction only?
- No, definitely not focused on instruction! I’m envisioning that we would cover all areas of library assessment - collection analysis, satisfaction surveys, occupancy analysis, etc.
- Angela: The November Viz and Tell is not going to be just assessment of visualization instruction but library assessment writ really broadly and we actually have, I’ve written up a short blurb of what I intend the November session to cover, so that’ll get posted on our website as soon as possible.
How much overlap do the different projects have? Are they completely separate projects or are they inter-related/inter-working?
- Justin: They do, they overlap a lot and really are sort of in the… you know, kind of crosstalk with each other. We have a month… Andy and Angela and I meet once a week, but then all the Fellows, we meet once a month, and we’re sort of constantly, you know, giving feedback, working on each other’s projects, and stuff like that, so they’re very interrelated and interworking and draw from each other and from the entire group.
Do you envision allowing people to add their own visulizations examples to the Teach Viz by Example?
- Absolutely! We will depend on it! We have a beta submission system already, but we thought we would wait to release it until we have done more user testing on the interface.
- Great! I noticed that two of yoru examples are ones that I use regularly in my teaching around data viz, so I would love to adapt the ones you have already shared to my teaching as well as to to add more from my stock to your list
Are the examples in the repository going to be licensed in any particular way or will it include copyrighted examples?
- We’ve definitely talked about licensing issues, and we may still have some work to do on that before we go fully live. Our hope is that we’ll be able to include copyrighted examples under some kind of fair use (commentary, educational, etc.).
- Yeah, sounds like a good approach. Great work here, Angela!
Are there training opportunities for interested information professionals/academics as part of the grant? And are these available to those outside the US?
- We are exploring that. The original purpose of the grant was to support the development of instructional modules, so we do want to preserve resources for those because we think they will be the most helpful for the largest number of people. But we have started talking about mentorship opportunities. Basically, our main training will be delivering the workshops.
How did you come to 1 meeting a month frequency? Would it not be better to have 2 meetings a month? Ideally where one day is tuesday and then an other day, in case of scheduling conflicts.
- If this refers to the Viz and Tells, we chose one meeting a month just because of the other commitments of the organizers. While we’ll be rotating the hosting duties around, hosting will be a significant time commitment, and our grant organizers and fellows are already stretched pretty thin!
When discussing basic literacy around data visualization comprehension and development, I have noticed in my own work that students often turn to us for an introduction to base concepts. I saw on the website there is an interest in teaching critical literacies at an undergraduate level, have you all discusssed what visulisation ideas are most critically needed?
- Angela (in chat): Haha, yes! We brainstormed a Data Viz 101 for undergrads just a few weeks ago
- Angela: And I’ll just I’ll just a hundred percent that. Students and faculty, I think, have looked to the libraries to give people a good foundation in data visualization base concepts and we are… we have just started working on data viz 101 modules at different levels, and undergrad is one of those kind of core levels that we really want to support. And so hopefully we’ll be seeing more work on the data viz 101 module in the next coming months. But the critical literacies really do build on this framework that we’ve alluded to, that Andy mentioned. It’s called a… it’s the data visualization literacy framework, I think, and there’s a link in the slides. It’s got a really nice set of core threshold concepts that we think are important for visualization literacy, and so applying that to our instructional modules is part of the work we plan for the for the rest of the year and ongoing.
- Yes, it was the libraries! And yes, we see both students, faculty, as well as staff. I have had a lot of administrators reach out as well.
- Sorry, the link you were mentioning, was that the report from the august meeting?
- The link I mentioned was to the data visualization literacy framework, inside the slides from today (it’s the link that says “threshold concepts”).
Is there an intent to include modules/instruction on dealing with perceptual and cognitive biases in data visualization?
- The Data Viz 101 modules in development now will likely cover something about perception and distortion in visualization. I’m not sure about cognitive biases, but we can look into it!
Is the repo for the site private?
- no, our repos should be public
- what is the github.com link?
- https://github.com/visualizingthefuture
- And https://github.com/visualizingthefuture/visualizingthefuture.github.io
How do people submit pull requests?
- You can use the basic GitHub functionality to fork the repository, make a change, and then submit a pull request for us to merge that change into the main repository.

About Doing Visualization

Please let us know whether there are resources about standardized formats for organizing data to create network diagrams and whether there are any recommendations for free, easy-to-use and install tools that could be used with the standard format.
- For network visualization, I haven’t checked in a while, but I think Graph Commons was really doing interesting things and make network data easier to produce and use. Maybe also Flourish?
- Gephi is a free network visualization tool, too (or was free last time I checked)
- Yes, also Cytoscape, if you are looking for downloadable tools
- I have another tool I’ve liked, but will need to look it up
- Downloadable tools that are easy to install and use standard formats
- Some answers from chat: Gephi, Cytoscape, Graph Commons, Flourish. The basic formats may vary from tool to tool, but often they will support one spreadsheet of “node” information and one spreadsheet of “edge” information
- Thanks for the Graph Commons recommendation!
We are talking about data visualization, what specific data do you need to visualize and why?
- That really varies! I think every workshop might need slightly different data to visualize.
Would love some examples of using large data sets - perhaps a community of interest on that?
- It might be good to check out existing data science communities for work like that - maybe machine learning or programming language communities. The R community might be a place to look for shared work on large datasets - things like TidyTuesday.

Save the Date

Finally, we have two more public conversations on visualization in libraries scheduled (so far). Our new Viz and Tell series will be a way for the library visualization community to meet semi-regularly to focus attention on specific topics.

October Viz and Tell
Topic: Teaching Data Viz in Libraries
Date: Tuesday, October 20
Time: 2pm Eastern / 1pm Central / 11am Pacific

November Viz and Tell
Topic: Visualization in Library Assessment
Date: Tuesday, November 17
Time: 2pm Eastern / 1pm Central / 11am Pacific

Registration details forthcoming. Look for a post here an in a few weeks and an announcement on our mailing list.