 Hello everyone. I think we're going to go ahead and get started. If it's okay with everyone, we're going to not use the podium, but switch back and forth over here at the desk. My name is Lisa Johnston. I work at the University of Minnesota Libraries. I'm the Research Data Management and Curation Lead for our data repository there. And I am Wendy Kuzlowski. I am a data curation specialist at the Cornell University Library. I also coordinate our campus-wide effort for research data management services. We are both a part of the data curation network, which is what we're hoping to talk to you a little bit about today. Introduce you to it if you don't know what the network is and give you some updates on what we've been doing in the last year or so. Sorry, we'll get the technology figured out here. So to give you an idea of what we hope to engage you in in the next hour or so, like I said, introduce you to what the DCN is. And at the same time, kind of introduce you to why or talk a bit about why we think data curation is important in this bigger picture of data management services. And then we have four main things that we want to talk about that the DCN has been focusing on over the last year or so. And then we hope to leave time at the end to get your input and ideas on this idea of collaborative support for research data management. Okay. So first, what is the data curation network? We are a collaboration of 14 institutions representing non-profit and academic data repositories that actively support research data sharing. We launched in 2016 with funding from the Alfred P. Sloan Foundation and have really, in the last year actually, moved into a sustainable, member-driven organization. Our mission for the data curation network is to really curate data ethically and in ways that make them fair. And we do this through advancing curation best practices, through offering training to develop the data curation profession. And as an organization, we really want to grow sustainably and responsibly to really best serve our members. Before we go into how we're doing all that, we're just going to pause here to take a moment to establish why data is so important to the scholarly communications landscape. And you really can think about this as two big problems, data sharing on the one hand and research rigor and data reuse on the other. Data are critical for reproducibility, increasing public trust and transparency, enabling new discoveries through reuse, reanalysis, or extracting, adding and building off of existing data. This example is a really good example of data sharing. This is just a few days ago from the CDC showing the number of COVID sequences that have been shared just here in the U.S., actually. There's a number of other repositories during the similar work. Almost 2 million COVID sequences that are available for public access use and analysis. On the other hand, there are still barriers. So even with something as large and impacting as COVID, where we all agree that open access to this research and the data behind it is essential, this is a recent analysis that looked at preprints and used text mining to identify markers for open data and open code. And the authors are still finding that the prevalence of that open data is pretty low. So looking at archive, they showed about 13 percent of the sample, only 13 percent of the sample had open data or open code markers. And that went up a little bit for bio-archive and then back down to about 15 percent or 12 percent for SOC and Med Archive. So we're really showing that even though we agree that data sharing is really important, there still must be barriers preventing this from happening successfully. And I will note that the authors, Collins and Alexander, also mentioned that this is substantially higher indicators of data sharing than pre-pandemic. So there are numerous challenges and they're listed out, a few of them, here on both sides of the equation, both for sharing data, really thinking about what data do I share, what repositories might I use, what are the privacy challenges or sensitive challenges for sharing that data. And then how do I reuse that data? Can I understand it? Is it reproducible? Can I even find it? So all of these issues are things that data curators, data librarians, data stewards, there's lots of ways to call us but we are all working collectively to try to really bridge this gap between researchers trying to share their data and others trying to reuse and test the rigor of that data. We do this through data curation and we try to facilitate reuse and long-term access to data. In particular, data curators try to be the first users of that data. As we're working with it in our repositories, we're working with our researchers to make it accessible. We will check the data for missing files, for missing documentation, we'll look at it through the lens of privacy issues and we're doing this work in collaboration with the researcher. And what I propose is that we also should be doing this in collaboration with each other and that's really what led us to launch the data curation network. Today, we're a community of about 40 plus data curators who work together through interest groups, through developing educational resources called data curation primers, creating opportunities to learn from our peer organizations and just really be a platform to share expertise, tools, tips and tricks for data curation. So now let's shift to some of the highlights of what we've been working on particularly over the last year. So the data curation network really does strive to enable ethical and fair sharing of data and we recognize that most data repositories are receiving a wide range of different data types. This is a variety of different file formats. You're receiving different code types, you're receiving different software packages. You are also seeing a lot of diversity in the type of discipline that the data are generated from. That is a lot for one institution to handle and my institution at least can't hire the expertise needed to cover all of those formats to really do curation that expert data level. So the data curation network creates a shared staffing platform to really allow for data sets to be matched with the experts that have that specialty in that data format, in that discipline and we really work across our institution to provide that broader layer of data curation support. We do this by really trying to seamlessly introduce the data curation network into local repositories or your repositories' existing workflows. So all of the repositories that are partnered in the data curation network all have their own ingest mechanisms, they run their own data repository platforms, they provide their own storage, they do their own appraisal and selection, that's all locally happening. When it comes to the actual review and termination of the fitness of that data and recommendations for how you might want to really augment that data for findability, discovery or reuse, that's where the data curation network can step in. And we will match that up with a curator across our network. They will make recommendations and then send those recommendations back to the local repository. So this really happens as a seamless microservice, if you will, within your repository stack and it integrates across all of our different repository platforms. One way we do this is by training our curators in a standard protocol. We call these the curated steps, conveniently enough. And these really allow us to provide consistent treatment to data across many different formats. And very particularly, we always want to include the researcher in that process. And that's the request step there, where we're looping in that researcher to some of the findings that we've made around this particular data set. And we work with them to implement those recommendations and then move that back into the repository. Now, one of the key things that we wanted to talk about that we really focused on in the last year is really thinking about how to look at our work through a racial justice lens and better incorporate inclusive and equitable data practices into our work. In this last year, we were fortunate enough to work with Dr. Faye Cobb Patton, who is a computer science researcher at NC State, and who worked with us to facilitate engagements with our network to identify key areas where we as curators could really incorporate diversity, equity, inclusion principles into the workflow that we have. So we evaluated our curate steps. And we really uncovered the fact that there are many unwritten steps that we as curators take when we evaluate data. And we do a lot of that work kind of in the background. And we really wanted to try to surface the ethical considerations that we as curators are making, particularly when it comes to sharing data about our human participants. So, for example, curators at our repositories, when we receive data that has been derived from human participants, we will request the consent form for what that participant agreed to, how was it described to those participants how the data might be shared. Occasionally, and probably more often than I would like, the participant form actually doesn't say a lot of clear things about how the data will be shared. And in fact, sometimes it actually suggests that the data will not be shared, and that it will be kept confidential forever and ever. So we really need to try to push back as curators to really recommend that the researcher needs to either go back and re-consent all of those individuals, give them a better awareness of how that data is going to be shared. And really, the researchers want to do the right thing here. They're trying to make decisions based on what their funder and their journals are requiring them to do. But also, we need to update our practices from the beginning when we're working with human data, human participant data. So, we took a look at our steps, and we also looked at a number of wonderful peer examples that really brought in a lot of the ethical components that we found missing in our current workflow. We looked at the fairness, accountability, transparency, and ethics in AI work fate. We also looked at the CARE principles, which really looks at how to handle indigenous data. We looked at the ACM's fairness, accountability, and transparency work, or FACT. And then finally, the wonderful Urban Institutes principles for advancing equitable data practice. All of this really led us to rethink our curation workflow and really explicitly build in the ethical components that we as curators say that we do. And we talk about it in our training, but we need to really write that down and be very clear and explicit about how we are evaluating data for sharing and the individuals involved and the communities that are impacted through that sharing process. So, this is all work that a subgroup of us has been taking a look at. And we're going to be hopefully releasing the updated curate steps in January. And that is actually really important to us because of my next update, which is really building out our community of practice. And for those of you who might not be aware, the DCN or the data curation network is unique in the fact that we really want to extend our training beyond just our curators within the network. We really want to open that up to really develop the broader profession of data curation. And so one of the ways we do that is through a two-day workshop that we've hosted with funding from the IMLS. We offered three out of four of the planned workshops right before COVID and reached over 80 participants who came to the sessions, had active opportunities to work with data, work with the curation process. And one of the things that was an outcome of each of these sessions is that we got people together for a capstone project where they would create what's called a data curation primer. They would work in teams of two to five individuals working over a six-month period to really research and develop out what are the specific steps that you might take for a particular format. And you can see all of the outputs from these workshops on the slide here. So we've we've developed about 24 of these resources that are really helping curators jumpstart that process if they might not be familiar with a particular data type. What we're hoping to do with this next project is to take that idea of matching a group together, a group with specialized expertise together with a problem and taking that a little bit broader in the research data lifecycle. So something that we're working on with Ithaca SNR is this idea of a data communities workshop. Ithaca put out this wonderful report in 2019 describing data communities as these fluid and informal networks of researchers who share and reuse a certain type of data. Most but not all of these data communities are facilitated through a website or an online repository. And a key thing is that data communities are not the same thing as a discipline. They're really working around a shared problem. One example of a data community might be FlyBase, for example. So the group that really comes around a particular genome of one species and really work on that together. COVID is another really great example. What the research that Ithaca did showed is that these communities really need help building out or identifying existing repository infrastructures. They're looking for technical and policy advice on metadata, vocabularies, preservation, privacy. They want guidance and advocacy for their financial sustainability. And most importantly, they're really just trying to reach other researchers in their community. And we really felt like the data curation network and data curators more generally can help with a lot of those problems. So we got an NSF grant to use the workshop model to really try to incubate data communities. And through an application process, we have selected 14 different data communities to work with. That'll be represented by researchers from that community to come together. We're holding this workshop in February of 22. To come together and match them with data curation experts in that area, either due to their expertise in the domain or file types they're working with, or some of the other challenging issues that that particular data community is looking at. So we're really going to hope that our shared staffing model for data curation and workflows for data repositories can extend itself throughout the research data lifecycle. So Lisa has already mentioned one research project. And while at its core, the data curation network is around curation. It really provides, because of its collaborative nature across these multiple institutions, a really unique platform for us to explore some of these ideas of interest in research projects that we have. So there are a couple more of these that we would like to talk to you about. And this one kind of is twofold. We wanted to take a look at, you know, we as outward facing librarians think curation is really important. It's very valuable to us. But do we really have a concept of how other stakeholders in this area feel about curation? So we did a couple of projects around assessing the value of curation from two different perspectives. First from repositories outside the data curation network, and then from those who are depositing content into the repository. So we're going to start with the side on the left here. And this would not be the right screen to switch. So this project looked at four key research questions. What level of curation are data repositories providing at this snapshot in time? And of those things that they do, what do they feel are the most valuable? And then what kind of impacts is that actually having downstream on the data? And how important is that to the larger environment around data sharing? And then this is not this there's not a cost to there is definitely a cost to data curation. So does the effort that we put into doing all this data curation? Is it worthwhile as the benefits outweigh that effort? So to get at that first question around levels of curation, we had to define what it is that we were talking about. So we are all speaking the same language. So we defined for the purposes of this survey, four levels of curation, of course, it could be that you're not doing any curation whatsoever. That would be what we refer to as level zero. And the level one curation is at the record level, working our way up through file level curation where the curators would open up the documents or open up the package, the data set package review the files and perform file formats, transformations if necessary to be sure that they're have they're accessible. One layer up from that level three would be making sure that the documentation is adequate and allows for reuse and understanding of the data. And then level four curation is where the curators are actually going in opening up the files, running code if it's there, making sure that the files interact in the way that the researcher expects them to in the platform that they expect them to do so. The survey was designed and intended to collect responses from US institutions. We adopted a non probabilistic sampling technique. So we knew that only research data institutions with data repositories were likely to respond. But we recruited participants from multiple list services and email lists. Due to that, we realized that there is a risk of bias and some data skew due to higher levels of curation for those who are likely to respond. And we recognize that this is a limitation when we go to generalize our results. The survey was open to collect responses for three weeks during January of this past year. So we had over 120 responses in total. We did, however, because of the listservs that we sent to got some responses from international repositories. The data that we are analyzing excluded those responses, which left us with 95 respondents in the United States. The data analysis that we're presenting is conducted based on overall responses, not at the individual repository level, because we had multiple responses from some repositories. Most of the respondents were repository staff. As you can see, we also had some responses from directors and a few from actual data depositors and repository users. They're mostly associated with disciplinary and institutional repositories. And like I said, because it did not limit the number, the survey didn't limit the number of responses from an individual repository. Places like Dryad and ICPSR and a few other really large repositories did have multiple responses. Taking that all out, there are 59 unique responses. And the responses were spread out over 23 states within the US, both some with core trust seal certification and 11 actual DCN member institutions. So we're going to present here just a few of those results. Looking at that first question that we addressed wondering what level of curation is actually happening within data repositories in the US. This chart shows the different levels of curations performed according to the responses. This question did allow for multiple responses. They're not necessarily, even though we developed those one through four, they're not necessarily hierarchical in their levels for the curation actions. But most often curation is being done at the record and documentation levels. You can see over 75% of the respondents had levels of curation at that. Responded at that they were doing that level of curation. We went on then to consider how often they're actually doing this. We can say we offer this as a service. We have the ability to do this, but for the data sets that actually come into our repositories, how often are we doing it? So we looked at the frequency with which those different levels of curation were actually being performed. And we can see that basically all levels, levels 124 are actually performed most or at least half of the time for 70% of the participants that responded. So it seems that most of the repositories in our sample are fairly committed to these curation actions and efforts. And they're doing them with the majority of the data sets that come in to their repositories. So we also presented the survey respondents with a list of curation actions. There are 30 curation actions that we presented to them and asked them to specifically identify the frequency with which each of those actions was being performed. And as you can see, there's a lot here. You don't really need to be able to read all the words. But if you look at those orange boxes, you can see that the most common curation actions were happening over 90% of the time around checking for duplicate files and review of the metadata and documentation for both accuracy and quality and completeness. On the other end, we see that closer to 30, 35% of the repositories were only performing, regularly performing the action of actually going in and editing the data for quality and accuracy. And so those things are happening much more rarely than some of the other ones. And we also took a look, remember, we had respondents from both disciplinary and institutional and generalist repositories. And then we did see a statistical difference between curation actions, using, you know, means comparison of means using Matt went to you. The results could be explained by the fact that these disciplinary repositories have a different set of skills that are much deeper than those being offered by institutional repositories. Remember, these are not all responses from DCN network repositories. So these are just how the individual repositories are performing these actions. So some differences between institutional and disciplinary repositories specifically in these areas. We also presented the survey respondents with a statement that data curation adds value to the data sharing process, and that data curation outweighs the effort and cost of data sharing. And the majority of respondents did agree with this statement. According to them, data curation impacts primarily on the ability for people to find, understand, use, access and preserve the data. And those are the most value added actions taken by the repository. So we hope to have a paper coming out about this, and it's going to go a lot more into some of the lessons learned and takeaways. But I think it's really interesting that there's a vast majority of the repositories, the data repositories currently in existence that are providing this kind of data level of curation. That's a lot of work. And despite that amount of work and investment that they're giving to their repositories and their curation actions, they believe that it adds value specifically around the idea of finding, understanding and using the data. Again, there are some differences between disciplinary and institutional repositories. However, we do acknowledge that this was a survey around the perception of the value that their work does on the data sets in their repositories. And we're not actually sure how that reflects around attitudes and behaviors outside those who responded. Also it was really interesting to see that when we had responses within a single repository, that individual perceptions about the practices that were done at that repositories didn't always match. So it's kind of interesting to think would be another project, more money. What would be how you could get a more authoritative answer around this, around the actual practices within a repository that are going on? So look for those in future iterations of the project. So I mentioned the other side of this was that we wanted to take a look at what the users of our data repositories actually feel about these curation actions that we're requesting them to take. So the second research project related to this, actually I should mention, a subset of the data curation network universities participated in an end user, what we refer to as an end user survey. And we sent out an 11 question survey between April and June of this year to anyone who had submitted data sets to our repositories who are doing curation between January of 2019 and March of 2021. And you can see here that while responses varied slightly between institutions, we had great response rate of 40%. Before I came to work in libraries and had never done a survey in my life, I would have said that 40% was an awful response rate. But I've learned that actually is a pretty good response rate. So we were pleased with that. And that represents 227 responses across those six institutions. So again, just a couple of interesting pieces that came out of these this work. Can't see my own numbers there. Almost 90% of the responses, either strongly or somewhat agreed that they were satisfied with the curatorial review that was received on the data sets that they submitted. And of those, the vast majority of data sets that had that went through the curation process had changes made to them. And they felt that because of that, that they were more confident in sharing their data and they're more comfortable, comfortable and confident in sharing their data. And because of that, not as we can't say that because of that, but that another question that we presented with them with was, do they felt did they feel that the repository added value to this whole process of data sharing and it was overwhelmingly positive over 95% said that they either agreed or strongly agreed that this process added value to the whole the curation process added value to the data sharing process as a whole. We also in that question around what value was added had some free text responses. And there was this is just a subset of some of the great things we heard back. But you can see in that middle one that they really felt for example that even though that they had thought about this a lot that the process itself was a still a learning opportunity. What is it that others might need to know and should be included in the metadata in order to increase the utility of their data set for the long term. This was repeated over and over again in the free text responses that this concept that having an outside set of eyes outside of the research project itself, look at the documentation to make it more understandable was extremely valuable. So getting at that last question so we realize that they all are they people agreed that it added value to the creation process and was it worth the effort? This is kind of really getting down to the nitty gritty. And again, over 90% strongly or somewhat agreed that even though this added an extra step to the data sharing process that the curation was definitely worth the effort. And again, we had some free text opportunities in this question as well. And we had a lot of free text responses over 100 people gave free text responses. And again, my process of learning how to do surveys I realized that free text is not something you always get even though it seems like an easy thing. And I think that this middle one again is really interesting. This it gets at this idea that these actions that this outside set of eyes these expert curators are able to make the process more comfortable for them. And that they would use the service again. So continuing on on the sum of the work that the DCN is doing as far as research, we also have recently received an NSF eager grant to look at look a little bit deeper into what the actual practices and costs are around data sharing. So this grant was awarded to the Association of Research Libraries, Cynthia Hudson Vitale is the PI in the project. And again, a subset of the data creation network institutions are involved in this work, all bringing perspectives from those institutions, which we hope to be able to to express as something useful outside of just those in the data creation network. And we acknowledge this idea of looking at specifically practices and and and costs especially has been looked at before. This isn't something brand new. People have tried to look at this before. And so part of this project also involves an advisory board from with including members of other institutions outside of the data creation network as well as Koger AU and APLU. So for this project, we're looking at four three main research questions. So for those people doing funded research, where are they sharing their data and what kind of metadata and of what quality is the metadata that is going out with those shared research data sets? And how are they figuring out and making decisions around where they want to share and what content they're sharing with their with their data sets? And then what is the cost to the institution? What different pieces come into play when evaluating the the actual cost of the the public access to share data sharing? So in order to get at those questions, we have kind of a three step plan, which is presented here very linearly. And as we are working on the project, we're discovering that it's not going to be linear whatsoever. But it's a nice visual. We want to take a look at using metadata, public metadata records, assess data repository use within and all this work is within five specific disciplines that are listed here. We want to then also perform a retrospective study of the actual data practices that faculty are using to develop service and infrastructure based functional models for how public access is actually taking place on our campuses. And then finally, we're going to use all that information, do some survey groups or interest groups and conduct some surveys to collect financial information on expenses related to public access, and hopefully be able to pilot and test existing financial models for this public access. All of this should be then relatable outside of I mentioned the institutions that are participating and have it be useful as a way to for other institutions to answer the same question, the same questions. And what we hope this work will produce is information and where researchers are sharing their data, which may allow for creation of workflows to support and fund those practices on our campuses. And by building those functional models, creating and documenting case studies and collecting costing data, we hope this work will provide new information for making data informed decisions for what the researchers actually want for services and infrastructure. It will uncover knowledge about decision making processes for research data sharing if they're willing to share out what it is that they go through in this decision making process and allow us to therefore create a public body of knowledge for practices and costs associated with public access to research data across a beginning set of disciplines and domains. All right. The fourth highlight that we wanted to share with you is really how are we really structurally able to do a lot of this work and what is the data creation network as an organization. So I'm just going to touch a little bit on some of the models that we put into place to try to achieve sustainability for this kind of collaboration, this multi institutional collaboration. So as I mentioned at the beginning the data creation network launched in 2016 with planning and implementation grants from the Sloan Foundation. We actually have kind of a go live moment in January 1st of 2019 where we really started that shared curation model. We got all of the technology and training in place to just kick off that shared model for curation. And then over the last three years you know we've really added additional members in a very controlled way. We really wanted to try to scale this up very slowly and intentionally so as not to grow too big too soon. So we went from six partners starting out in 2016 to today where we've got 14 partners and this year 2021 is really where we've been able to transition into a membership model. And that model relies on a new governance structure that we put into place. So we've got a governance board that includes representation from each of the sustaining member institutions. So each institution has a representative and this group really thinks about the strategic directions of the data curation network really focuses in on what are our values as a community. How do we really live those values and consider you know projects and and work that really aligns with that. We also have a small executive team that meets weekly to really address any questions that come up on a more you know immediate basis. And we've got a wonderful advisory board consisting of individuals mostly from from the member institutions leadership at each of those institutions to really help us figure out where to go next and what what is on the horizon that we need to be paying attention to and work with. We the membership model itself is is really reflective of all of the operating costs in the data curation network. We've got one full-time assistant director salary. We host an annual event the All Hands meeting. We actually worked in travel costs into our membership model to really at least for this year wisely protect some of those travel funds so that our attendees would be able to to still go to some of these events. Because that in-person networking really has been a key component for the data curation network. This is all fiscally home at the University of Minnesota. So we've been the lead PI on the grant since the beginning. And that transition staying at the University of Minnesota was sort of a logical choice at this point. We really didn't want to lose all the wonderful services that are in place at at university and try to go it alone at least not at this moment. And then one other structural element here represented on the slide is just the fact that we really recognize the different types of work that we're doing. Some of it fits in sort of a committee type model. Maybe ongoing curation work that's happening. A lot of the structural issues around how do we scale up our educational programs. That's all happening via committee. But we also recognize that there is room for others to join us. And that's really we've developed these special interest groups. And that's where a lot of the the research is happening where we've got groups coming together to talk about issues around handling big data. You know how do you really move data around at your institution. How can we learn from other peers around this particular problem. And so that's where our DCN community is open to others to come and join us and continue those conversations. And that's been a really fantastic way to really keep in touch with our attendees from the workshops or or just really reach out beyond just the DCN membership. I mentioned we do have a membership model in place. So this actually went live this year. We onboarded those 14 members at the sustainer level. That's where all of our members are right now. So they're each paying a ten thousand dollar per year annual fee plus in kind. And that in kind is really the engine that drives the data curation networks shared staffing model. So we do have the curators contributing between one to five percent up to one to five percent of their time. We actually are very careful not to exceed that time. We track that very dedicated. We track that in a tool called Jira. We also are experimenting with other other ways to get involved for institutions. You know we obviously are currently represented by a lot of larger institutions and maybe more well resourced or more well staffed for curation. What would that look like for an institution that maybe has just one person doing data or zero people working on data curation and really wanting to you know become a part of this network. So we are beta testing you know a different member tier that would allow for you know working with the network but maybe not with the same kind of staffing commitment or or even maybe the curation aspects at all. The ambassador tier is really where we would like to host more workshops in the future after we're doing in-person workshops again. So really getting an opportunity to take some of our training to different regions and really upscale around data curation best practices. And we've tinkered around with this idea of a sponsorship tier as well that might help cover some of the cost for all hands meeting. And again we when we are doing those again that would be where we're going with that. So just to kind of wrap up here of what can you do to get involved. Well you could join as a member institution. We have been doing an application process so far. We are you know still trying to grow slowly and really thinking about bringing on institutions that help us fill some of the curation gaps that we have. So we really are seeing a lot of data sets coming into the network that have significant code components and a variety of different software languages. So really looking for institutions that could provide some of that expertise that we might be lacking or that we might be over utilizing in some you know core individuals time. We also want to really extend our training as I mentioned. So inviting the DCN to come in and provide our specialized data curation workshop setting up opportunities for the creation of additional data curation primers. That's really benefiting you know the entire community. And then those special interest groups are open to all. We've got a list of them on our website with different contact information for all the different topics that we're looking at right now. So some takeaways here. We think you know building off of what we've been doing the past several years that when it comes to providing data support and data services collaboration across institutions is really key. These problems are not unique to your institution or your researchers. These are problems that we all have and we can really approach them as as a collective and be more successful that way. Shared outcomes really drive any collaboration. And for us that means better research data data that is more understandable more usable and ethically shared. And we recognize that this happens and this benefits us regardless of institutional origin. So data being shared at Cornell is going to benefit my institution and my researchers as well. So that's why we can all kind of get on board and collectively do this together. We know however that there are definitely some barriers to making all this happen. And that's what we wanted to do today is just kind of open it up to not only your questions about our talk. But what are some additional barriers that might be preventing us from really collaborating at the radical scale that we're trying to do. And what would you know what would we really need to do to help make that possible. So with that we will close out our talk and invite any any questions. Thank you so much. That is a full time job. We have a full time assistant director that really is taking a look at how often we are asking our curators to curate those data sets. So we use JIRA as I mentioned. We ask all of our curators to log the time that they spend on a particular data set. And so far our capacity we've stayed within our capacity. So as I mentioned we've got about 40 data curators. Any given time you know we'll see like four or five data sets go through the data curation network for shared curation. Our institutions are handling a lot of the data curation still locally. They're not you know sending everything to the data curation network they're sending well the hard ones to the data curation network. The other thing we do is we talk to our curators a lot. We do annual partner check ins with all of our members and we sit down and talk to the curators about their feelings about that workload how you know how often are they getting tapped is that okay. Oh and one other key thing at the end of the workflow we use JIRA to track the curation assignment. We asked them at the end of every curation assignment how did that go. Was this the kind of data that you'd like to receive in the future. Did this match your skills and expertise. Did you feel confident curating this data set. We've really found that if you give me a data set that fits within like the file formats that I'm extremely expert at and the discipline I love and care about I'm going to have an awesome time and I'm going to be excited to see the next data set. If I get a data set that comes to me that is of a file format I've never heard of or maybe a software type I know I don't have the expertise. I'm going to be a little bit more cranky about that. So we really try to stick within what people are really feeling expert in. I also might add from the other perspective around load bearing the other benefit of the network is that we don't necessarily institutions aren't submitting just because they're lacking the skills to curate a specific data set. It also helps us load bear locally if we just have too much coming in at once and like I'm perfectly capable of curating this data set except that I'm also curating six others because we each have limited capacity as well. So it isn't just about the skill set but being having the flexibility to as needed push those data sets that we might be able to handle otherwise to the network and they have more capacity there to handle those. So that has also been really beneficial and thus far hasn't impacted the capacity of the network itself. Thank you for all the University of Nevada Las Vegas. Can you speak a little bit to how you work with research offices and maybe information technology offices within a university not just the I'm assuming that's mostly libraries but could you talk a little bit more about the institutional commitment. Thank you. Actually one of our working groups is is really looking at that kind of campus advocacy piece of how we communicate this this curation type work with others on our campuses not just the researchers but those in IT and the research offices. I'd say it's happening pretty similarly at most of our institutions where there is often a campus wide group thinking about research data or cyber infrastructure issues. You are right. Most of the the members involved are in the libraries are based in the libraries but not all of us and that's actually one of the exciting things that we really want to try to expand on is really working with groups like the Michael J. Fox Foundation for example who are really bringing a perspective that is you know different from from our library perspectives and hearing about their data curation challenges which are quite similar exactly the same and really expanding the types of partners and individuals involved in this work. So I guess it's to say that there's that that is happening at all of our campuses. We do try to compare how we are going about it and compare the different types of campus infrastructures for collaborating with IT and we want to bring in those voices as much as we can. I think one thing that we're learning in being able to do that together is trying to figure out the best way to communicate to those different groups. It's very easy for those of us who kind of in a very insular way look at data curation and throughout these words and activities and say we're going to curate your data. But if we want to get the message that this is a service that could be advertised through the office of research for example what are what is the terminology that represents or that resonates with that group and with the people who they talk to which could be the researchers but it could also be for example research administrators who may not have as clear of an understanding of the whole process of what goes on during curation. So being able to talk amongst ourselves about that how to communicate what it is that we're doing with these different groups who might need to hear it in a different voice or using different terminology. I wonder if you could talk a little more about your relationship to specialized repositories or specialized platforms that are repositories plus offering other analysis sorts of services. I'm thinking in particular of some of the things that the National Institute of Health is funding these days. Those are often pretty well funded and have a support package around them of staff associated with that. How do the members of your network relate to researchers who are or perhaps should but don't know about these platforms? So taking the latter half of that question I think that of the at least the academic institutions that are a part of the network, all of us have the practice of that idea of appraisal of whether the data set is even appropriate for going into our repository. And the very first step is often to have that conversation with the researchers and make sure that there isn't a place that is more appropriate or being used more commonly with their peers. So I think that that happens outside of the network itself and we've a lot of us have educational information locally to express that. As far as the DCNs interaction you may be able to speak even more to this than I but one of the other things that we offer the community of curators is a we make an effort to speak regularly with these groups. So we have had you know portage come and work very closely and explain their practices. We have had conversations with the National Library of Medicine. We have talked with the Woods Hole group frictionless frictionless data, trying to talk and learn from all these different repository platforms and efforts and groups that are working on similar efforts. What is it that that we can learn not just as a network but that the curators can bring home and implement pieces of at our home institutions. And I'll just add that there is a gap you know between when people in our role are reviewing a data management plan for example and the data management plan might describe you know utilizing a specialized data repository or disciplinary data repository. And you know we're advocating for those repositories at that moment and then you know three or four years passes and they actually share the data. And we're often not brought back into that process. We could be we're an underutilized resource there where we could be working with them to curate the data for really any data repository and preparing that for going to those data repositories. But we're just often not looped back into that process. So I think it's a real it's an interesting problem of how you could utilize this group of experts who really do have a vested interest in making sure that data reaches a trusted destination and is well used. But really we only see that data if it's coming to our local repository. So I think we need to find better ways to insert ourselves into that process. All right. Well if there are no other questions we'll certainly be hanging out if you want to talk to us later. But thank you so much for coming and we really appreciate your interest.