 Hello, and welcome to our session. I am Danielle Cooper, and with me is my colleague Dylan Rudiger. And today, we will be sharing about it because research portfolio related to data support services and researchers' needs related to that. So we'll start off by sharing a bit about the data support landscape as it's currently configured, and then do a deep dive into the researcher's perspective on their support needs based on our most recently published report. And given the nature of this room, our ambitions around discussion may be tempered, but we certainly have discussion questions for you that we would be eager to hear about. So we begin with the question, what does the research data support landscape look like, and how can it be optimized to meet the needs of researchers? This is not a new question in many ways, but we would argue that this is a really good time to be asking this question. It's certainly a moment where universities across the US are thinking about their data support services in a different way than they were even five years ago. We've gone beyond the phase of just being enthusiastic and being additive about the services that we want to provide to our researchers and really thinking more about how they can be coordinated, centralized, made more sophisticated. I'd also like to acknowledge that it's a really great time to be having this conversation at CNI. Our presentation is part of a longer conversation that's been sustained starting with the executive roundtables over the past few weeks. There was a really incredible session this morning offered through DryEd about how they're evolving their membership and service model. We had the previous panel just before this session on data science support services. And tomorrow, I'm certainly looking forward to the update on collections as data and on fare for the US. So it's quite a robust conversation that's happening. So in terms of how we've been approaching this issue at Ithaca SNR, and I'm imagining that for most of us here, we have a pretty good familiarity with Ithaca SNR. But for any of those who may not be familiar, we are a not-for-profit research organization that specializes on issues related to information and technology practices. So data support services and the needs of researchers is an issue that we've been tracking in a number of ways for a number of years. To give you some examples of the reports that we've put out recently, we did a inventory of research data services in the US. And that was published in 2020. We just came up with a report on researchers' needs that relate to big data. And these are the two reports we'll be focusing on for the remainder of our time together today. But in addition to that, we did an issue brief on research cores. We have an ongoing project on data communities that's funded by NSF. We are doing a project where we'll have the findings published later this year on teaching support services related to data in the social sciences. And we also have a forthcoming report that will focus on data practices in the humanities. In addition to that, we do work at Ithaca where we track the practices and needs of researchers longitudinally. So we have a national faculty survey that we have been launching triannially since the year 2000. And that includes questions about data practices. We have a series of projects on how researchers do their work in the data is collected qualitatively. And we do it in different disciplines. And over a number of years, a number of the disciplines have been data-focused fields that we have findings related to that. Our data communities work is actually ongoing. It goes beyond the NSF project. We do a series where we spotlight different data communities and their practices. And we also do work focused on the research enterprise and the role of the senior research officer, including a report that was put out in 2020. And of course, data is a major priority for that stakeholder group. So just to give you a sense of the sheer scale by which we've been doing work on this topic, we have done well over 1,000 interviews with researchers where we've engaged with them about their data practices. And we typically do this work collaboratively. We've worked with over 300 institutions with their libraries specifically and their librarians to conduct interviews with researchers. So today, we're focusing on two studies in particular, where we did mixed methods research on data support needs and users' practices in this field. I'll start with the inventory, which is from 2020. And this is simply just a recap to get us started before the main event of this presentation. So in 2020, we did a web-based inventory of 120 US colleges and universities across Carnegie classifications to quantify the number and type of data services that were being offered. This was a report that was published in 2020. And we are very eager to do an update of this inventory and have plans to do that later this year. Something that I really want to emphasize here about the importance of taking this approach to understanding how data services are evolving in the US is that if we were to do a survey, so reach out to an individual at every higher education institution, have to pick one person and ask them about the scope of the services at their institution, we'd likely get a very uneven response. And that's not just because people don't like answering surveys necessarily, but because it's really hard to get a handle of all of the different data services that are in any one institution. It's hard enough to ask an individual to try to characterize that when what we found over time is that typically there isn't that level of understanding in most institutions. So it's also very hard, of course, to basically get every institution's information. It's quite labor-intensive. So we go with the approach of doing a representative sample of US institutions. And we've developed a very systematic methodology for then evaluating what services are offered as displayed publicly through their websites. Some highlights from when we did this original study in 2020 and the kinds of trends we're tracking in this area over time. The first is that there is quite a difference in terms of the type of institution and the volume of services that they're offering. We see that at our ones, there's typically about seven to eight data services being offered. Whereas when you go down to R2, that drops precipitously to more of the two slash three range. Then you have small liberal arts colleges where it's about the one to two range. But given the difference between an R2 and a small liberal arts college, that gives you a pretty clear picture of the level of resources that can be invested. So this is something that we definitely think is important to track over time. And quantity and quality are not the same thing, but it is really staggering to see that difference. The second piece, which is especially important at CNI, is to recognize how essential libraries are to the offering of data support services. What we found was a cross institution type. The library is typically at the center of what is being offered. Of course, there are other decentralized offerings that you find quite a bit, typically in bioinformatics or related to statistics. But the library is always part of the picture. And this is something that really needs to continue to be championed and recognized because the libraries have really stepped up here in terms of what they're offering. So now I'm gonna turn it over to Dylan because there's one way we can tell this story, which is what institutions are currently doing. And then there's what the researchers actually want and need. Thanks, Danielle. So as Danielle has already mentioned, our inventory gave us a pretty good idea of what the landscape of offerings that are out there looks like, what kinds of workshops, trainings are being offered and by whom. It was less effective by design at giving us a picture of how researchers were engaging with and using those services, how widely used those services were, and also how well they were meeting researchers' needs. It's also worth pointing out, particularly in light of the panel that we were just in on radical data support services that perhaps, even though the research we did on the inventory is pretty recent, we may see some usage trends changing due to COVID and the transition to virtual offerings. But either way, we wanted to learn more about the experience of researchers who were using these services and also to understand a little more about how researchers were engaged in data-intensive research to try to get a sense of how they conducted their research and what kind of support needs they articulated as emerging at the top of their agenda so that we could get a sense of how to align the offering, or how universities could align the offerings that they're making with what researchers on their campuses are telling them that they want to do. To do this, we launched the Supporting Big Data Cohort. This was a cohort of 23 universities in the United States. Each of them put together a team of librarians who conducted interviews on their campuses with anywhere between eight to about 15 researchers working actively in big data research. Together, that gave us well over 200 interviews. These were hour-long, semi-structured interviews about research practices and support needs. From approximately 100 different departments, so a very wide angle view on big data practices across disciplines ranging from the usual suspects that you might expect would be heavily engaged in data-intensive research, physics, bioinformatics, and so forth, to humanities fields, social science fields, business schools, law schools, high schools, really the whole breadth of contemporary modern universities. Last December, we published findings from that report, or from that project, and I'm gonna talk briefly about five key finding areas that we highlighted in the report that we published in December. Those are disciplinarity and interdisciplinarity, complex data and managing it, workflows, sharing knowledges, and how instructors articulated their support needs and practices. I'll try to keep things really high level, and I encourage you to read the report if you wanna dive in a little more deeply, but I'll just say on the issue of disciplinarity, that big data research is commonly a collective endeavor, and it almost invariably involves interdisciplinary domains of knowledge, unsurprisingly, but not all disciplines can contribute to this equally and participate equally in data-intensive research. There are different disciplinary incentive structures, different methodologies, different research cultures, and of course, different levels of access to resources and to funding that play a pretty profound role in shaping researchers' ability to contribute to big data research and to practice data-intensive research, and that also affects the level at which they need support varies considerably from discipline to discipline. It's also worth pointing out that our project interviews turned up pretty considerable evidence of some tensions between traditional disciplinary ways of knowing, and methodologies drawn from the computer sciences and data sciences. Some of this was kind of a friendly rivalry. Some of it was maybe academic turf battles, but some of it was really well-informed and well-founded sense that traditional disciplinary ways of thinking and seeing the world that had long lineages were at risk of being kind of threatened by what some researchers perceived as kind of an overriding method of drawing from the data sciences that was kind of taking space that once might have been the domain of an anthropologist or a humanist or even a physicist in many fields. And this raised some tensions between those kind of disciplinary-based ways of knowing and more interdisciplinary practices that are really required to make data-intensive research worked. On the topic of managing complex data, we found that very few researchers really struggled to access data, although discoverability was a significant challenge. Many researchers found that with some diligent searching they were able to find data that they could use. Perhaps most interestingly, one of the things that our interviews turned out was that many big data researchers are now relying heavily on secondary data, which they preferred in many situations as cheaper to obtain than data that they generated themselves. Well, there are certainly fields that this is an exception and there are many researchers who continue to generate their own research data. We found a great many researchers who really either preferred to use secondary data or relied on a mixture of data that they generated themselves, supplemented in some ratio or another, by other people's data. And so we're starting to see pretty widespread evidence of data reuse, although it's worth pointing out when we get to sharing that there may be an unevenness with which people contribute to the data pool that they like to draw data out of. And so that problem remains, but researchers are reusing data. When they do so, they're doing it in very complicated ways. It will surprise no one in this room, I'm sure, to know that big data research is almost always a collective endeavor that involves many hands and lots of work and that organizing the workflow is perhaps the most challenging part of much data intensive research. One thing that I'd really like to highlight here, and it resonates with much of what we were hearing in Joel's panel recently, is how important students are to the work that's going on on campus. Labs are conducting much big data research on campus, and inside labs, you have undergraduates, you have graduate students, you have postdocs, you have staff, you have faculty of all ranks. Really, you have the full spectrum of communities that comprise the university who are working on data projects. And undergraduate students and graduate students are playing very important roles, not only in cleaning and wrangling data, but in generating data and in coding, and in the case of graduate students in particular, in interpreting the results. This raises the question of whether or not we need to begin framing questions about research data support as much as pedagogical issues as research issues and thinking about the pedagogical connections between undergraduates, graduate students, and the data work that's going on campus. That seemed to be something that was a very common thread throughout the project findings, and it was true in basically any of the fields that were organized around labs, and even many humanities fields and other fields that don't have a strong lab culture are kind of adopting that workflow as they turn to data projects of a larger scale. So it suggests the need for the kind of undergraduate and graduate student-oriented data support services that we heard about in the other room just a few minutes ago. On the topic of sharing knowledge, what we found was that, broadly speaking, researchers were pretty committed to the idea of openly sharing research outputs. But that they tended to do so in ways that reflected a broader spectrum of practices than we might think of as kind of formal, fair data sharing. And in fact, we walked away from this pretty convinced that sharing is really a spectrum, that it ranges from highly informal practices of emailing your buddy with some data all the way up to publishing data inside of an institutional repository, and that there were researchers felt for some good reasons and other reasons that we might find a little more questionable that all of those practices had their place. This risks, of course, creating data that kind of moves in closed circuits and through networks that some people may have uneven access to, but it also created opportunities to share data with relatively minimal work to prepare that data with other research teams who were well positioned already to understand and work with data that may not have metadata or other documentation that was sufficient for audiences that were less specialized to be able to make sense of. We also found that many researchers felt that they were under pressure to share and to archive and preserve more data than they really felt was appropriate. This varied quite a bit from researcher to researcher, and it reflects the amount of labor that goes into preserving data, but there were a number of people who really felt concern that the kind of pull towards open sharing of data was leaving them feeling like they had to share data that they didn't really want to share, whether because it was derivative, excuse me, derivative, thank you. Once you get tongue tied, you can't get your way out of it, low quality or gathered from sources that had ethical reasons that made them reluctant to share it. Finally, on the matter of support and training, the big takeaway here was that for immediate kind of just-in-time training and support needs, many researchers are quite content to rely on Google, YouTube, web tutorials of various kinds, which they see as appropriate for the problems that confront them in their labs while they're doing their research. They run into a coding problem, they don't know how to solve, they just want to go on Google and figure out how to solve that problem, and then move on. However, they did seem to think that, well, they might be reluctant to use the kind of more formal workshops and trainings that libraries and other units are offering, that they were happy to recommend students go there, and they expressed a real concern for conceptual and foundational types of training offerings. Many of them felt that if they needed to learn how to solve a problem, that they could learn how to solve that problem, how to make a piece of code run smoothly, but they felt that what was falling through the gaps was their ability to comprehend the full spectrum of what they were doing. Blacksbox issues related to software, related to ethics, related to what happens when you mix methodologies from different disciplines together, loomed large in their sense of what they were missing out of, and they really appreciated opportunities when they were provided by libraries and others to speak in depth with their colleagues about these kind of trickier, thornier conceptual issues that they didn't otherwise have organic ways to make space for in the course of their research. So, given all that, the question that we wanted to pose today was how can universities align their support offerings to the things that researchers are saying that they need? And the answer to this is obviously gonna vary quite a bit. We know there are a lot of institutions, many of them probably in this room, who are doing really good jobs already at providing the support that researchers need. There are others who have further to go, but our report illustrates several key areas where libraries and university units might think as they look forward to the future. The core challenges are that libraries and other services have is the need for diversified and highly individualized services. There's a wide range of methods and tools that are in use, and the users that are using them come to it with a wide range of competencies. Many also perceive their needs as very idiosyncratic or rooted in disciplinary ways of thinking that make it hard to transfer to other contexts. This may be a communication problem, it may be a real issue in some cases, but it's something that many researchers feel as though their problem is very specific to them. It is also the case that on many campuses, data support services are siloed, decentralized, and even if the university has a pretty good grasp on how to map those resources across campus, those maps are not necessarily legible to users who have trouble knowing where to go with what kinds of problems and even seeing what's available across complex university webpages. And finally, there continue to be problems with antiquated senses of what libraries are. Many researchers that we interviewed were very quick to suggest that libraries were just places to get a book or an article and saw them as essentially a kind of retrieval service. And really, even in a few cases, ask somewhat hostile questions actually about like, wait, you're at the library, why do you care about my research practices and support needs? And so there's work to be done as there has probably always been work to be done, making sure that faculty understand the kinds of services that are being offered in libraries. In terms of aligning services to those needs and in the face of those challenges, a good first step, and this is, we found this and also the recent ARL and Carl study on a similar topic that concluded along much the same lines, that there's a real need to make sure that you're mapping and coordinating resources across campus to assess data services, to coordinate them across units and to try to find ways to minimize duplication and to amplify the value of the services that you're offering. It may also be the case that it's worthwhile thinking about targeting training opportunities to disciplines that are less likely to have practitioners who are already come to the table knowing are or Python, particularly at the kind of introductory level of workshops and trainings. There may be more space in some fields than others to attract and to market those services. Another thing that we found that would be really useful at least according to researchers is creating spaces and forums for people from different disciplines and backgrounds to share and network together. Over and over again, we heard that researchers really, because their work was so interdisciplinary, researchers really valued the opportunity to get to know colleagues from other fields, to learn about methods and tools that they weren't already aware of and to just talk to people in other disciplines so that they can conceptualize projects and to learn where their interests overlapped. This seems like a promising direction in particular for getting around some of the places where researchers were prone to thinking that their problems were a little more idiosyncratic than they possibly were. This is something that we've seen come up in our recent NSF project on data communities where we brought together STEM and social science researchers from a wide variety of fields who entered the workshop really thinking of their problems as pretty idiosyncratic and individualized and over the course of conversation with each other over a couple of days, walked away with a greater understanding that well the fields that they might need for metadata or documentation may differ that the conceptual underpinnings of them shared a lot of common ground. And so there's spaces for libraries in particular as kind of neutral ground on many campuses to host to these kinds of interdisciplinary convenings and forums and networking events and it's a useful opportunity. And finally, perhaps most difficultly, we were just talking about this before the break, demands on data storage and accessibility. Data storage needs are just growing and growing and growing. And they look posed to continue to do so at a pretty exponential rate, particularly as more disciplines come to engage in data intensive research as kind of a normative way of proceeding. And so there's a real need for universities to think about how they're going to prepare to be able to store the data that's already accumulating and that is going to come in future years. So those are some of the high level findings from the Big Data Report. In terms of next steps, we're beginning to think about where we go from here and I'll just briefly highlight two things before we move to the discussion. The first, as Danielle already alluded to, is our forthcoming research data services assessment cohort project. We're still in the very early stages of conceptualizing exactly what we're going to do here but we are building off of both our research and the ARL-CARL report, which suggests that many campuses have ways to go to map their resources, to catalog those resources and to create resources that users at various career stages will find legible to help them navigate the support services that are being offered. So later this year we'll be putting together a cohort of libraries who will conduct that kind of assessment mapping, cataloging, exercise as a group of co-learners and who will conduct interviews with people on their campus who are using resources across units to see how well aligned their offerings are with researchers' current and future needs. And the idea with that is that it will allow participants to use that evidence to create or evolve their data service strategies and also to create these maps and resources that will be useful to a variety of users on their campuses. And finally we'll be publishing some new material later this spring. First up is an issue brief that we're working on now on data sharing practices in the humanities with an eye towards understanding not only what humanists could learn from debates and conversations around open science but also what humanists might be able to contribute to those debates themselves. And also later this spring a report also based on a cohort project with a large number of libraries on teaching with data in the social sciences which really reflects our sense that when we think about data support services it's as much a pedagogical issue as a research issue and that the two are overlapping in increasing ways. We're at the discussion phase here. Danielle I'm gonna turn things back over to you. You're certainly welcome to come up to the mics and ask any questions of us you like. However, if one feels more like sharing about what's going on at their institution we also welcome that as well. Here are some things we're especially interested in learning about or talking about more recognizing of course that this room is not really a room for discussion but these are the things that we always love to ask institutions about. The first is simply what are the biggest challenges your institution is facing while it's developing its data support services and perhaps simple was not the right thing to say there because I wouldn't argue it's that simple at all. We'd like to learn more about the extent to which your institution is interested in or engaged with data support services that are cross institutional and we ask that especially because it's such huge ad narration for example of how the data curation network is working and we'd love to hear more about your perspectives on those kinds of models. And then finally, what information about data service trends and user needs would be most useful to your institution especially because as Dylan was sharing we will be doing a cohort project where we will work with the group of institutions to assess their data services. But as we do that project we'll also be updating the inventory we did of data services across the US for as a benchmarking tool. We're planning, this is kind of a quasi longitudinal effort so we'll try to keep a certain number of the fields we used in the original study stable but we especially are interested to hear if there's certain kinds of data points or metrics you'd like to see tracked over time in this area. But again, if you wanna ask us any other questions but we presented that is also welcome to you do not feel beholden to these questions at all. Good afternoon, thank you for your presentation. I was not able to join the earlier sessions that you mentioned but I wonder how much you engaged with vice provost or research office staff as an example. As collaborators on a campus I mentioned this because I have found them to be very open and willing to work with my librarians. In fact they invite my librarian to do the workshops for the researchers and the other question is graduate students seem to be the main audience for data services support workshops. They flock on a Friday afternoon and state after five o'clock for these sessions. We find them to be much more recipient than the faculty who are sort of stayed in their way of thinking. So maybe I'll start with the senior research officer and then Dylan you can move on to the grad student component. So for the studies that we discussed today are focused on we did not engage with the senior research officers because the first one is an inventory of services that we did looking at websites. So didn't talk to anybody. And then the second study that Dylan was focusing on around big data supports. In that study we interviewed the researchers themselves. However I agree with you entirely about what you've observed at your own campus because we did another study that we didn't discuss today recently that was focused on senior research officers. And it is abundantly clear that data support is an important priority for them. I think you could argue that the extent to which they see the library playing a role in that is variable. I wouldn't argue that it's consistent at all. I think you could also do additional research there to bolster that because our original senior research officer study was not focused exclusively on data support. It was on a series of topics that are of relevance to those individuals. However, I would definitely think that this is the big question for libraries as they continue to explore how to evolve their data support services is making sure that it aligns and is legible to those in the VPR or senior research officer type role. There is certainly a lot of area there for growth but the way that libraries typically talk about it with other libraries is not necessarily consistent with how those and other administrative roles would think about it. And I think there's a lot to ask about in terms of simple things like, well not simple, but brass packs kind of things around budgetary lines and who's paying for this because in my opinion, the most successful institutions are those where they've really aligned the data services across units. It's not just the library, it's also IT, it's also research computing and there's certainly a unique role that the library can play in that. I shout out to Susan Ivy who earlier today talked about NC State and they're doing really interesting things there and they've really figured out a way where the library can play a role in an almost customer service like way be kind of the place that people are willing to go to. But within all of this is the question of how the staffing works across units and who pays for it and that's where you really need to have the VPR on board. Over to you Dylan. Sure and on the question of graduate students, my basic response would be absolutely yes. It was very clear from the interviews and makes very intuitive sense that in research labs, graduate students are workhorses who are doing a lot of the actual work with the data. At various stages they are also very involved in writing the code. Several of the researchers who we interviewed were willing to acknowledge that the major intellectual contributions were really coming from the graduate students that they were providing a framework and a supervisory role and that the people who are most well positioned to catch mistakes and to really understand the data where the graduate students in their lab. We heard in the panel next door, most recently the one that you weren't at, that graduate students are in fact, fairly frequent users of these services at least on some campuses and that many of them are coming to graduate school feeling as though they're already behind the curve in their understanding of the tools that they're expected to know. They also seem to be perhaps more willing to show up to workshops and trainings that are scheduled for times that maybe they don't have control over. So I do think that's a really important space and I think it's part of how we can start to frame these questions of support services as pedagogical things and also where we might start seeing collaborations between libraries and research computing centers and places like the Center for Teaching and Learning and other units on campus that are more oriented towards the teaching mission, thinking about ways that they could collaborate to bring this into graduate seminars, undergraduate seminars and to further think kind of the boundary lines between research, teaching and how those things connect inside of data intensive labs. I wanna complicate the picture a little bit further though because what I found especially fascinating in the session this morning that Dryad did was the fact that the institutions seem to observe that their power users are faculty and postdocs and that's because Dryad focuses their offering on the concept of publishing which is something that's a bit later in one's research career. Ah ha ha, I mean a faculty member can be relatively junior but hopefully you understand what I mean here. And we could have probably a philosophical debate about the extent to which Dryad is a platform versus a service and certainly this session is about data support services but I was also really impressed to hear from Dryad member institutions that it was their curation services that were incredibly valuable that Dryad is offering curation so that one doesn't have to do that kind of staffing within the library and this is something that is valuable to researchers and postdocs. So of course one strategy is to make sure that your data support services are valuable to graduate students or students because there's certainly a lot of demand there and the session that was just prior to this one even emphasized the importance of undergrads and how much the need is there and those coming from high school who don't necessarily have the skills yet but then on the other side there is something to be said about services that are valuable to researchers especially since they're often reticent to see the library is having the kinds of supporter skills necessary for them whereas it seems like at least with Dryad they've hit on something in that respect. Hi, Scott Walter, San Diego State University. Building on that, that struck me as something that fits into what information about trends and needs would be useful. For us, we're what I might think of as sort of a high level, high performing R2 with aspirations to make that leap into the higher classification but we're not structured in terms of the staffing and some of the investments and some of the things that you're mentioning which is probably why if I were to take the inventory we would be that 2.8 rather than that 6.7. So when you were talking about Dryad as a platform versus a service, right? Services are delivered through platforms. You make a choice, you take this institutional repository or that institutional repository and part of it is what sort of shared services might be available if you're not able to staff it up entirely yourself. So that gets to my question about information might be useful. Collaborative approaches. Are there tools out there? Are there collaborative efforts out there that might allow a greater range of institutions in a world where most of us are not going to be able to fully staff up a full scale research data management service to collaborate? And that could be everything from shared platforms that people contribute to to the development of high quality Zoom based workshops that could be deployed across different institutions and probably any number of other things that I'm not thinking about. But to me that's really a critical question. If I don't have the ability to fully develop a research data management department even though my university is moving strongly in that direction, where can I invest collaboratively with others to bring those resources to my folks? Yeah, I think that's a really good point, Scott. And I'm sure that DryEd would be very intrat, I feel like I'm like the DryEd, like a, I don't know, referral service shiller that there's no, I was given no payback but I feel like they'd really appreciate your perspective because that was definitely something that came up quite a bit in the session this morning. To even take that one step further, it's not just about staffing up but going back to the earlier comment around where is the senior research officer, the VPR, but also I'd say the provost or the president right now, it's the reality is that it's incredibly difficult to compete with other industries, to hire staff with any sort of know-how that is IT or data related. It was hard before the pandemic and it got harder during the pandemic and it's gonna continue to be harder now that we have the issues surrounding inflation. It's incredibly difficult for universities to compete for this kind of talent so I can imagine that any sort of effort where there's a centralization and taking the localized staffing pressures off the table make a huge difference. Thank you for that overview, Maggie Farrell, University of Nevada, Las Vegas. So following a little bit up on that and thinking about the challenges, one of the challenges is positions. And repurposing existing positions and so I don't know if you've explored how funding structures work because this is really moving libraries in a dynamic, great way but it's also an additive to our existing work so could you speak a little bit about what you found in terms of creating the expertise within the libraries to do it and again, following up a bit on Scott's notion about collaboration across the university so I have a libraries might be working with that expertise with other departments in order to lift up the entire university. Thanks Maggie and there's limits to the concept of doing an inventory approach to mapping services and I think you hit your head on one of them. You can't tell from a website the extent to which staff were successfully repurposed for other roles so to be very candid we wouldn't have that answer in that respect but Dylan, you sound like you had something to say I've been dominating, okay. So but I agree with you entirely and I think you've also hit the nail on the head on this other theme that we keep on returning to which is where our leadership outside the library thinking about libraries and the staffing models is definitely something that we've heard through our various channels is an issue for leaders. They don't necessarily understand the current staffing models. They don't understand why staff can't be repurposed. I think it's a wicked problem. I mean, here we have, it's almost like a building the plane while you're trying to fly it because the universities themselves are creating these concepts of professionalization around data and so we don't necessarily have those cohorts of people who could then staff the equivalent services for them so I'm sorry I don't have an answer but I think you bring up a really good point. The only thing I would add to that is just to emphasize the possibilities for coordinating things across the university, right? Like libraries clearly play an important role now and hopefully will continue to do so but it may be the case that there are staffing options and expertise in other places on campus that can be leveraged for some of these. There are statistics departments that offer certain kinds of services. There are programs that train undergraduate students or graduate students to be data support ambassadors either to peers or to students and faculty and so there may be opportunities that exist outside of the library to coordinate things that take at least a little bit of the edge off those very real problems that you're raising. I'd like to bump up the student model even further because what strikes me is that actually historically libraries have been really fantastic employers of students and there are some examples of institutions that are already leveraging students really, really successfully to do certain components of these support services and I definitely think this is something that the libraries should continue to focus on. It's there's for so many students it's just so much easier to go to your peers when you need help anyways and again the library has clearly demonstrated over time that it can be a really successful employer of students. Question or comments? Hi, Sandy Thompson, University of Houston Libraries and curious to know based on the last several kinds of questions if you all have been looking at what makes for a successful service like are there particular metrics or are there sets of trends you're noticing that begin to make this kind of work more sellable to administrators or more meaningful to other stakeholders to what I imagine continue to do the work in the future? Yeah, I think that's a tricky question because success is kind of perspectival right? Like who are the services successful for and certain kinds of metrics about say attendance may be a good way for the hosting organization to do a rough and ready idea of how successful it's being what we're hearing from researchers about the services that they value the most are ones that are very difficult to deliver comprehensively what researchers tend to value above everything else are one on one consultations relationships with librarians or other staff that if not necessarily long term are at least face to face and intimate in a way so that the consulting staff or faculty member is someone who has a real sense of the project and of the specific kinds of questions that the researchers are asking those are obviously among the most labor intensive kind of services that you can offer and so that's kind of like you know not necessarily happy news but those are the things in our interviews that when they're available researchers really really appreciate there are also campuses that have invested again often in through departments actually statistics departments in particular in building up staff or faculty resources you can provide really high level expertise in statistical analysis or in data science again those look a little bit like the consultive relationships but they may be less intensive but they leverage real high levels of expertise the services that tended to get to be perceived by researchers as the least successful overall frankly we're the intro to Java intro to our intro to Python workshops which many faculty and researchers walked away from feeling like they didn't get very much out of now had we talked to more graduate students and postdocs and we did interview some of them that may look very different but that again points to my initial comment about success depending on whose success you're talking about and I just like to add that I mean this question's infinitely fascinating and I definitely think it's important for institutions to come up with ways of assessing systematically the extent that their students or their researchers are satisfied with the services or what they need that's one side of it and then you have the administrators those are not the same metrics by any extent what the story you tell your administrators could be different but what's especially fascinating to me right now in terms of the concept of success and what an administrator perspective on this is is that unlike how library services are traditionally funded there's definitely an expectation that this is the kind of thing that could be involving cost recovery or and you know that you see that with the more specialized or decentralized services like through statistics or bioinformatics I definitely anticipate that that expectation will carry over with services in general in this area I could not predict how possible it is to do that but I do think that that would be an expectation and then there's also these kinds of more nebulous questions that are hard to map on to data support services concretely such as retention how do you talk about this in terms of retention that seems like a really hard baseline but I wouldn't be surprised for an administrator to ask that question we're technically at time but I would like to take this last question but please do not feel beholden if you have to leave I think there's something else starting relatively soon oh but I wanted the last word okay go go go Carol Wattley Caltech I wanted to add something that I think fits nicely with what you were just saying because we were talking about like how do you make the case to campus administration about needs for more resources in this area I think it's really dependent on the audience that you're talking to around this right so I actually report to the vice provost for research and so I haven't had to make the case for this except I mean like he gets it but at Caltech the conversation around this is more about how the library can provide support to researchers so that they aren't distracted from the real business of the research and at previous institutions the way to go about it was to do this kind of benchmarking thing that you're talking about right and compare yourself to your peers or to your aspirational groups about what services they offer how well they staff them and that sort of thing and sort of deciding where you want to be an A level and where it's okay just to be average right It's a really helpful observation Any, are we done? Did this Come and find us after thank you for coming and let's keep the conversation going