 Great to see you all. Thank you so much for coming to our presentation, and thanks for your patience during our little technical difficulties We're really excited today to share more with you about our project on models for sustainable and inclusive Data science consulting and collaboration in higher education. It's a real mouthful So we'll start with a little introduction and overview of This and so who are we on Mara Blake? I'm the head of data and visualization services at the North Carolina State University Libraries Yeah, and I'm Emily Griffith. I'm an associate professor of the practice in our department of statistics and more Relevantly for this project the director of analytics consulting for our data science Academy at NC State And you might be asking yourself What is data science and consulting and collaboration and just so we're all on the same page for the presentation? we're really referring to advisory or possibly embedded support on data science and statistical tools techniques and methods and this might happen at any Stage in the life cycle of a project So today we'll go through the introduction and overview and you can see we'll talk a little bit about a local context a project We worked on what we're planning next and time permitting some questions and discussion So we'll start with sharing a little bit about our local context at North Carolina State University Yeah, so we have a long history of data science related programs and disciplines at NC State You can see it starts with operations research in 1970 and continues on through our most recent I think program the foundations of data science master's degree, but we wanted to show this slide to show there's a lot of activity at NC State There's also if you need support for your data science related project There are multiple avenues of support the Department of Statistics offers some Collaborations and also a class where graduate students will work with you at no charge For a semester the libraries have the data and visualization services department and the data science Academy has a collaborative consulting core And so we've coordinated these programs together to avoid duplicating effort or duplicating offerings intentionally or unintentionally We want to ensure seamless user access and nothing is more frustrating than being bounced from person to person If you have what feels like should be a pretty simple question And we also want to be sure that we're meeting the growing needs for data science support across campus And that includes touching on new disciplines as they're getting more involved new techniques as they're developed This is a constantly changing landscape So I'll share a little bit about what we're doing with data science in the libraries at NC State And I'm going to talk a little bit about my department data and visualization services and obviously the focus of today is consulting So that's what I'm going to talk most about But just to let you all know that Consulting is part of the suite of services offered by data and visualization services and that suite also includes workshops open to all our campus users course embedded instruction sessions on data science topics and Specialized lab spaces that have hardware and software that help support people access the tools they need to undertake data science work So we support everything from finding data Wrangling that data getting it prepared for analysis analyzing visualizing working in geographic information systems and accessing software And now we'll focus it down to consulting and how do we support data science consulting so in terms of staffing We have professional faculty librarians focused on data individualization. They support our users offer consultations as Well as a team of graduate student data science consultants where we typically employ 8 to 10 students who are offering Consulting to campus as well. We also have a staff specialist a data science support specialist who helps us run and manage our program And in terms of what that looks like for users and the support they can expect we're typically working around a 30 minute Consultation time might be a couple appointments building on each other But sort of that smaller term and really focused on advisory support and guidance So not production work and users might access that dropping into our lab spaces where our students physically staff the space or They might schedule and that's based on expertise So there'd be a little matchmaking on your question is about using Python. So you'll be matched with a data science consultant who Has expertise in that area and the consultations could happen in person or virtually via zoom So the data science Academy Consulting Corps only provides data science consulting We don't provide workshops or any sort of instruction modules like the libraries do And this program was designed to fill a specific gap in Analytics support that was uncovered at the University a few years ago When we did a pretty comprehensive assessment and found that small projects get help with the libraries and very large Interesting projects can find faculty to do some methods development with them And there's sort of this gap in the middle of Medium-sized projects that aren't interesting enough to do methods development, but are too big For some of the existing services on campus Because of that we're also focused on more specific areas of that research life cycle Specifically wrangling analyzing and some visualizing of data and we won't help you find data And we don't do software access Staffing wise we're similar in that we have a group of graduate students. We have five to six Maybe seven and then instead of a team of librarians. We have one faculty director and that's me So it's a much smaller service in scope But we offer up to about 20 hours of dedicated support per project And so we do not do long-term embedded partnerships. We can facilitate those We can be scheduled we're always scheduled We don't have a physical space and they can be in person or via zoom But our students will code for you. They will help you interpret your results design your experiment and do those sort of longer term communication Reliant projects and so We collaborate really closely Between our two groups again to make sure our services complement each other Just like you know, we're making sure we offer the 30 minute versus this median term We work together closely to create a shared intake process so users can contact both groups through one place and it's all Invisible to them. They just get the help they need and we sort it out on the back end. We Conduct a joint student hiring process so that we can make That a little smoother. We're making sure we're meeting regularly communicating with each other about things that might affect both our groups and We also share an interest in Working on projects and grants that might advance data science consulting as a subfield in an area of research to help support people like us working in this area to do it better and That led to the next thing that will share a little more with you about which is our project on Consulting models and so we were really fortunate that The Alford peace loan Foundation shared our interest in this work and supported a one-year project for To convene a cohort of experts working on data science and statistical consulting models You know bring it together a lot of types of programs types of institutions types of individuals and Looking at enhancing what is we find a limited literature and a limited body of work on the topic and Making it a little less accessible to approach this type of work at people's local institutions, and so we were really fortunate to get this and this is a shared project between the data science Academy the libraries and The academic data science Alliance sometimes abbreviated to at the also participated in this project with us So the goals of this project were really to look at the structure of how Data science consulting happens at organizations, so not how to do a data science consultation which many of our colleagues at other institutions Do really well and don't need help on but they don't know how to get it set up So that it's funded and they're staffed and hired to be in a position to offer that consulting with which they have expertise So looking at Administratively, where is this work located on a university campus? How is it funded? How is it staffed? What kinds of practices can we learn that we can share with each other versus what do we need to adapt to our local context? How do we make this work more sustainable more equitable and How can we disseminate findings about this to help support people like us working in this subfield? So we structured the project the way you can see here. We had virtual introduction slides and we had four Webinar based pre meetings around different topics and data science consulting that led up to a day one and a half day in-person workshop that we hosted at NC State in the spring of 2023 and Had plenary sessions as part of that a really wonderful keynote on equity and inclusion in data science consulting Which kind of primed our participants to divide up into writing groups? different topics in this field and Since that meeting they've been working asynchronously to to generate some content on that and we've been checking in with them virtually a synchronously We had a wonderful group of participants I think we're fortunate to have one or two here at this meeting. So that's really exciting and we had really a Mix of representation we cultivated this intentionally to bring people from different administrative locations like academic departments libraries standalone centers on campus different types of institutions participants with different roles in their individual programs and a mix of demographics although We do want to acknowledge that in this field. There's some some room to grow in terms of diversity and inclusion You can see the fine folks who participated in this project here. It was a wonderful group I don't think we could say enough good about it And we felt very fortunate that we got good feedback in return from our collaborators Who shared that the group was very engaging and also helped them? Learn things that they can apply to their local context, which was our goal. So that was really exciting We had some some takeaways that We're happy to let you know about this is preliminary of course but we really found the structure of the project successful the pre-meetings led to a very Dynamic and kind of engaged group of people when we finally got together in person We learned from our colleagues about the great work that they're doing and their programs that was really exciting And we learned that while there are some things that are generalizable There is a lot that has to be customized to local context and that we still see that continued need for more Research-based recommended practices in this area to really Enhance this type of work make it so other people can develop their own programs around data science consulting So and one thing I realized that we didn't include in that slide was that this workshop also led to sort of a spur of the moment I guess ish webinar Featuring two of our higher level administrators at NC State who held a webinar for participants about how to pitch Consulting programs to your own administration. So how do you come up with talking points? Who do you talk to? How do you find out who to talk to? And really how to craft that kind of proposal and that was at the request of some of the participants and it was really really wonderful So for our next steps and we're sort of going in three directions at the same time We're working on integrating our programs more tightly It's a real benefit to us as as the faculty and the librarians It's a benefit to the students to have that bigger cohort and I think we all are really enjoying working together as well We're also looking at the current student position types and hiring and seeing what we can streamline and what we can update based on Sort of the status on campus of students in the kinds of positions that they are looking for and that are competitive for them And then the last one is the impact of these emerging areas of data science And that can be new methods new techniques new ideas Anything like that that's sort of coming and becoming a need so that we can make sure that we're ready to address those questions when we get them We also have a forthcoming special issue of the journal stat and it's focusing on models in Statistical and data science consulting and so the the real target for this or the initial target was the papers coming out of the workshop that we organized But we also have an open call and we've had a couple papers from that open call already And we know of more in the works if you have one Feel free to submit it. The call is open through the end of January and we can discuss extensions if needed but that issue will be coming out in 2024 and we're really excited to see Just the richness being added to the body of literature that exists about consulting structure in higher education We also have an additional grant Which just got funded again by the Alfred peace loan foundation. We're really excited and this one has two Two arms the first one is supporting conference presentations and professional meeting attendance to share This kind of work and we're hoping to make an impact not just in Like the conference on statistical practice, which has some of this right but in meetings like CNI In other meetings across multiple disciplines including biostatistics libraries Basically anything if you're interested will have an application form and we're also working We have a group of co-authors identified who are going to write a publication. That's a manual style publication looking at leveraging graduate student employees in a consulting program and that gets at that academic mission of a lot of our programs to Educate and to train that next generation of researchers and scientists And so we would really like to thank the NC State Libraries the data science Academy and the Alfred peace loan foundation for their financial and Other support and the academic data science Alliance for partnering with us and all the participants Of course in our models in data science consulting workshop who just did truly exciting and exceptional work And really I think reinvigorated our efforts in this program So now we have time for questions Hi there Joan lippencock CNI emerita Wonderful and interesting presentation. Thanks for all the detail. I'm interested in knowing two things and even more but What disciplines are you covering are both of your units covering a wide range of disciplines? Can you give me some idea of the types of disciplines that you're working with particularly and secondly is who's funding the data science Academy? I'm you know the library. I assume a lot of that funding is internal You may be getting some special university funds, but who's funding the data science Academy? Thanks I can address the data science Academy question first because that's probably the simplest The academies are a new initiative at NC State and they're funded by the provost office And the idea is that they're sort of big cross-cutting initiatives that need to be truly touching every college every department And so they should live at that university level. So that's who's funding the data science Academy We have I think a five or six year commitment right now and are working on making ourselves indispensable As far as the disciplines we have computer science several types of engineers statistics Natural resources natural resources Quantitative public policy economics social sciences so a Wide wide array of students It's really interdisciplinary in terms of both our users So for the libraries when we track our data We see users across every college and division With emphasis and this is the emphasis of our university engineering the sciences Natural resources Being our biggest users, but we also hire that same way to reflect that interdisciplinary nature So across both our groups we hire students who are representing a really wide range So we we also employ and get that interdisciplinary experience that way Hi, I'm Cara Watley Caltech and one of the things that struck me in Your slides early on were the reliance of both of your programs on graduate student employees And I'm familiar enough with the library to know that you will have also full-time Staff consultants and librarians there And then I saw that you have plans for a publication on graduate students positions like this and I I'm kind of interested in that In that model and I'm wondering I Think that I would have a problem Staffing a service relying on graduate students and I'm imagining that there are other people who are in a similar Situation at their institutions and so I'm just wondering a little bit about your thinking on it and if any of the other groups that you were working with have sort of plans to transition to more I Don't know like permanent full-time jobs for these sorts of things or this is not a fully formed question So anyway, go with it where you would thank you That's a great question and that certainly Occupied a lot of our conversations at at the meeting and the work the workshop meeting among our participants and we had participants representing Programs who were staffed exclusively by staff scientists so full-time staff data scientists all the way to Groups who are using undergraduate students, which is something that you know, we'd maybe be a little apprehensive about pursuing So it's it's a really good question and we've encountered some challenges hiring and retaining graduate students because the students with data science skills are In in big demand across campus and in the private sector even so But it's I think something really near and dear for both of our programs where we Love working with the graduate students where they bring in really fresh skills really fresh perspectives versus our staff in some cases not the experience in communication or You know some of that interdisciplinary work But we feel like we get so much out of our work with them And they get so much out of their work with us and sort of this experiential learning kind of experience where they Work outside their disciplines. They develop really great communication skills And I think they report to both of us that that is something that helps them be really competitive in the job market And that's something that's in turn helped us recruit and retain students when possible Do you think I think so and there's also in statistics at least there's a pretty long history of working with graduate students to do Statistical consulting and so I think the leap for me was not graduate students But it was sort of trusting Engineers to do this kind of thing right because I'd only ever worked with statisticians when I started working with the Data Science Academy the breadth is is Really different from what I was used to but they all do like a really good job And I think we work pretty hard to have very open communication So if they're stuck on something they're not worried about talking to us about it because that's the real trick You don't want them hiding that they're stuck I'm curious about How you demonstrate return on investment or what methods you're using to Go about making the case that there are benefits being accrued Sure So that's a really good question We're working on we have some student positions that have been written into grants with a similar model And so that's one argument for more permanent funding We keep track of how many appointments we have how many projects we work on So that's another sort of if you can show kind of a need It's not that expensive to hire graduate students really And so that's another part of the argument and so we're kind of hitting multiple points at the same time So it's a good educational experience for the students It's good for the university to be on the cutting edge of this people are interested in giving us money In addition hopefully to that that more permanent commitment from the university and then Yeah, that there's a need and maybe some publications and things that can come out of it as well It's very similar in the library's context where we we track our usage and impact very closely where the data Visualization department. We're all about the data. So we're we're using that heavily and leaning on that to demonstrate our impact on campus John Peters Virginia Tech Following our colleague from Caltech I think it'd be really interesting to see the job postings y'all put up to get successful hires to For graduate students to help do this work That would be really cool to see or if that's something that y'all talked about postings for for higher students across the workshops That'd be really cool stuff to see Yeah, I that's a that's a great suggestion. Thank you so much for that I think we're happy to share a job posting and then the the manual that Emily described that will be that forthcoming publication will include Deeper information about crafting job descriptions to hire graduate students as well as like the advertising and interviewing process For this type of work. So we'll be we'll be More details on that will come out and we'll be having that available in the meantime I think we can definitely share that if there aren't any more questions. We would like to thank all of you for attending We really appreciate it