 Welcome. Again, my name is Marseille Curie-Lidu and I'm here with Sara Ann Murphy, who's coordinator of assessment at Ohio State University, Jeremy Bueller, who's the assessment librarian at the University of British Columbia, Rachel Lwela, who's the assessment librarian at the University of Massachusetts, Amherst, and myself. And here's a slide. I'll come back to this slide. So you have at least a visual of our faces as we all engage into this, but I hope to be a discussion. We do have a large group of people attending, so we will ask you to submit your questions through the chat box. So please type your questions. We have received some questions already. And in addition to covering those questions, Sara did a little bit more of a slide deck for us based on some of the questions we received in her session. I hope all of you attending have watched the three webcasts that each one of our individual presenters did, showing how they are using Tableau. Jeremy, Sara, and Rachel, have you watched the other two people's webcasts? Yes. They did. Indeed. Any questions? Let's start with asking you if you have any questions of the others. Rachel, anything you want? I've had the pleasure of talking to and working with Jeremy and Sara all along, so they've had lots of questions from me, but I don't know if I have any right now. That's fine. Anyone else wants to venture or has something to ask from each other before we go to what our participants have offered to us? Sara? I don't have any questions right now. And Jeremy? No questions. I was just delighted to see that what I already knew of both Sara and Rachel's work has developed even further as the presentations show, so that's great. Wonderful. The three of you have been working together for a long time. The way I became more familiar with your work is through the panel you did at the Library Assessment Conference in Seattle last August. And of course this work is evolving as your experience with Tableau is evolving. And we are now having a growing community, I hope. As we were preparing for these webcasts, I wanted to remind everybody of the resource that Rachel made me aware of. It was an article that appeared, the Gardner Report. It was the February 2015 Gardner Report. And it appeared right before our webcast series was starting. It's a report on business intelligence and analytics platforms. And it does have this chart there with the data they collect on the different analytics platforms, sort of grouping them into this quadrant based on two dimensions, the ability to execute and their completeness of the vision of that analytics platform. And you can see on the top right quadrant there, you have the leaders in this field in the business intelligence and analytics software field. And there are a number of software packages there, companies. Beyond Tableau, SAS, and SAP, and IBM is there, and Oracle is there, MicroStrategy, Microsoft, what's that, quick. But Tableau, it's sort of a little bit of an outlier up there in terms of the ability to execute, the ability to deploy it quickly. And the report does conclude that this is sort of the gold standard. And that's what sort of led our webcast series to focus on this product. But there are other products out there. So some of the questions we have received already from our participants has to do with the cost and the pricing models that you can deploy and pay. And we're fortunate because we have different pricing models amongst the three of you. Rachel, would you like to tell us a little bit more how your pricing model works for you? How much does it cost? How many people have access? Sure. And I would encourage folks who are interested in checking more specifically in terms of a quote with their rep, and they can contact Tableau and they'll get their sort of academic region rep who will be really helpful sorting that out. In fact, I spoke with our rep, Adam Putter, this afternoon. So we have academic pricing. So we have desktop professional, which is 1,500 the first year. So 1,500 your first year with a 300 annual maintenance fee. We now have two copies of that. So we have two staff members, myself and Jessica Adamick, who have the desktop professional, which is sort of the authoring tool. You can get Tableau personal for 750 and that's 150 annual fee, which just lets you work with flat files and you can only publish to the public server. So you can't save those locally. So that's one of the differences between the personal and the professional. So we have the professional because it lets us connect to multiple data sources, the database connections, and that's something that personal doesn't do either. Those are just flat files. So two copies of that, two people doing authoring. We publish much of our data to the free public server so that anyone can look at it. And there's no fee for that. But we also have our own library version of server which we run on our hardware. There are multiple ways of setting up the server licenses. So really this shouldn't limit how people think about it. But the minimum server configuration is for 10 licenses and it's 750 per login. And that's on your own equipment. And there's $150 ongoing annual cost with that. So you can also subscribe to Tableau Online, which is their cloud server where they hosted their hardware. And I think that's 500 per user per year. So that wasn't even, I'm not even sure that was an option when we purchased server here in our library, but that's sort of the alternate choice. And then there's a large sort of institutional six-figure implementation that is really for the whole institution or for systems, which are priced and configured separately, but significantly more expensive. And we do have institutional implementation in this group, right, Jeremy? And is that your Tableau? That's correct. And that UBC has an institutional license. We're still on the fence in that the library hasn't yet committed to participating in that because there's a cost-sharing component and we're still weighing the benefits of going that route. From what I can see, that seems to make sense. But I'm afraid I can't discuss the costing model for that other than to echo what Rachel said that it ends up that's distributed across an entire campus. That may be a more affordable option, but you're definitely dealing in overall with a much, much higher subscription cost. I'll just extend from there and talk a little bit about where we are right now with Tableau implementation at UBC Library, very similar to what Rachel described. We have, I believe it's now three Tableau desktop professional licenses. And I have found that for those of us, there are a few in the library who use tools, or formerly use tools like Excel or other data presentation tools quite heavily. Tableau desktop professional has been worthwhile investment in its own right just in order to support our own work. Never mind the publishing possibilities and then using Tableau public where possible. And if we do commit as a library to being part of the institutional foundation of Tableau, then the Tableau server is obviously a bonus for distributing that. But even on the data analysis side, I think Tableau desktop professional may be worth considering on its own even if you can't afford a server. Very good. And Sarah, how is the setup at Ohio State? Well at Ohio State, I am the only individual with a desktop server, I'm sorry, desktop professional license. I actually started with Tableau desktop personal. That was $750 with a $150 annual maintenance fee. And the nice thing about the annual maintenance fee is it includes all upgrades to the next version of Tableau. And it seems that since I've had Tableau, it upgrades to a new version almost every year. I was able when I had Tableau desktop personal to save my own files. But I now have Tableau desktop professionals so that I can interact with more than just Excel files and access files. I can connect directly to data sources such as Google Analytics, anything with an ODBC connection, etc. And so that is $1,500 a year. And I have been able to share desktop, share dashboards and workbooks using what's called the packaged workbook option in Tableau desktop professional. This allows you to extract all of your data and then you then share it in this alternative file format. And then anyone in our institution can open that file with the free Tableau reader. And so what we do is we, I upload files, Tableau packaged workbook files to like our cloud environment if it's appropriate. We use Box here on campus. And people can then download them onto their PC and open the file from their desktop. Or if it's something that needs to be on our internal servers, we'll just share it in the shared server space. And all they have as long as I have extracted all the files and placed everything in the one folder, they can open it from their desktop with Tableau reader. So we have made a conscious decision right now not to pursue Tableau server. Mainly because of the cost involved with having to pay for each individual login and the minimum 10 licenses. It just wasn't cost effective for us at this time. Yeah Nancy Turner actually answered her question because she was asking is it necessary to have the server in the library. So you can imagine a scenario where the server is in the institution and you share that server with some other unit or department, right? Well I would love to have server. I would love to have server because then you can embed your visuals on websites but keep your data private. I can't do that right now. I can embed visuals. I can embed dashboards on websites but it has to be publicly available data. So I have some interactive dashboards already on a couple of our libraries' websites but it's all publicly available data so I'm not worried about that. But I would like the ability to present some data sets but yet not have to share the data that's driving that data set with others. And are there any other units on campus you think that may be using Tableau and could potentially in the future be collaborators? We have a pretty active, well we're starting to have a pretty active Tableau users group here on campus and we're finding more and more people every month who are interested in using Tableau and we're trying to bring everyone together to learn from each other and also we're having this new feature where you can bring a dashboard, you're having trouble with making it work the way you want it to work and you can crowdsource how to fix it. That's been really, really helpful. So we also have a very active users group. Tableau users group has just been the Columbus Greater Metropolitan area. So I think I went to the first meeting years ago and like a couple years ago and we're like we fit around one table and now they're pretty big. That's interesting. And I like to go to the city-wide meetings just because I get ideas on how to do, how to put things, how to put dashboards together. It jogs my thinking. It helps my creative juice flowing. Melissa is asking if Rachel, would you please repeat the cost of the server, the number of license, and the maintenance fees? Okay. So the minimum if you have your own server hosted locally, you provide the hardware. A minimum of 10 licenses are required. That's a sort of baseline or 10 licenses. It's $750 per login and that's perpetual license so that's a one-time cost and that includes your $150 annual fee for the first year. So you pay $150 times 10 ongoing for a minimum of 10 licenses to start with. You can buy more too. And I think if you get high enough there may be other pricing but that's sort of the baseline pricing. Rachel, it's Jeremy here with just a question about this. When you talk about logins for this, am I correct in understanding that this is in addition to your desktop professional? So you have your authoring platforms, that's that other pricing model you described. And then the logins are just authenticated access to whatever it is that you publish on server. Is that correct? I think I followed you correctly. I use up one of my server licenses so I have to have one too. They're not in addition to their separate. So is it still possible with this setup that you can publish data on the server that is accessible to anyone who has the link and does not need a login? Or is everything in this model behind a login and using one of your 10 logins? Everything in this model is behind the server and requires one of the 10 logins. And they're, I think, set up, they're intended to be individual logins. So not like the reference department login. It's for a person in the reference department. So it can change if somebody leaves, you can transfer that license to somebody else. But it's not meant to be a generic license. It's meant to be a named license. Which is why we end up publishing a lot to public because we have more people, more than 10 people that we need to access our data. So we have to keep track of what we're good at sharing publicly versus what we need to keep on the server. This, in many ways, is a big challenge with data, right? You know, what do you release at what level? And there are different levels of disclosure and clearly that has implications on the tableau licenses. But Shian Brannon actually is asking a very interesting and provocative question. Has anybody considered making tableau available for student use on library computers? Their GIS librarian is considering this. Have you thought of tableau as a service for the student community? We just added tableau public to the build in one of our computer labs. But at Ohio State, we did that mainly so that we could have a class in that room. We haven't really started promoting it as a service here right now at this time. At UBC Library, we haven't considered that, and I appreciate the question. Tableau public seems like a viable way of doing that, of adding that into a build. I do know also that tableau provides free desktop licenses for students, as it just has until recently. And I don't know whether an arrangement could be made with them in a lab setting to have the more fully-fledged tableau installed on the desktop. I asked about that, and the answer was no. So that is why I put tableau public on the machines. And it's only in one room in one of our library computer labs right now. Because I was teaching a class for librarians in there. So it's there if anybody wants to use it, but we haven't really told anyone that it's there yet. I don't know if there's anyone who's from the University of Washington who might be able to chime in, but I think I remember reading that they were a school that did an institutional license, and part of that was with all students having free access that I think they were building in that kind of student support for youth as well as for publishing. So I was really interested to read about that at the time. But I'm not really sure of the details. It's a little hard to figure out who might be from the University of Washington because there is one attending, but if there is one who has actually logged in with the web interface and they can do the voice through the web interface, you can raise your hand and we will be more easily identifying you this way. If you have used the phone line, it's a little harder to identify which phone number corresponds to which name. So if from any of the participants who has this ability and would like to raise your hand if you have logged in through IP access and not through the phone, raise your hand. In the meantime, I think this brings us nicely to the second set of questions we keep receiving about Tableau. How do you train people? How much time does it take to learn it? I know it's an exciting piece of software, but it does have a learning curve. Sorry, you did mention one of the reasons you installed the software public Tableau in the lab was to teach your colleagues about how to use it. So do you want to tell us a little bit more about the learning curve and how much effort does it take to learn it? While Tableau does offer a number of really good video tutorials on their website that you can watch when you're starting to interact with the software and learn how to use it, I did find them a little challenging because they were designed for individuals who are in the business community. So it just took a little bit of translation of terms into how everything was done in profit and revenue. So I just changed my thinking to think about usage and other things so that the training would make a little more sense to me after a little bit of practice that got a little better. But I've been using Tableau since spring of 2012 now. I will say that it's pretty easy to make a simple visualization right away. It's a little more challenging to make an elegant visualization that takes a little time and a little bit of learning and a really good interactive visualization with higher level filters. I'm still learning about that and I might be learning for quite some time because I've got to learn how to do some calculations and other things first. But it really depends on what you want to do, how you want to use the software, how much time you're able to put into it like anything. It gets easier with the more you practice. It has a bit of cadence so as you use it more you start to realize, oh yes, I need to look for this on my Mark's card. I need to look for this item over here on my column's shelf. At first that makes no sense and it sounds like I'm talking gibberish but after a while it does work. The reason we gave an instruction session here at OSU is because our data management librarian received feedback that our librarians were interested in learning more about data visualization so we teamed up and she took about 45 minutes of the program to talk about some data visualization tools and then I took another 45 minutes to talk about Tableau because we've learned over time that data visualization means different things to different constituent populations, etc. We actually did this last week. There were only about seven people at the session but we did download Tableau Public onto our machines in the computer lab and it was part of the build. The nice thing about having Tableau Public on the build was that I just put all the files for this program up in a box and over time one of my colleagues came into the room and we just added them to the box file and then they could download the Tableau files that I provided and immediately open them on their machines. That was nice. I did create some job aids for them and we walked through how to put together one type of dashboard. I think it was word clouds and text tables. Thank you. How much time is it taking you to feel that you are learning different elements? I'm never sure how to answer that question because in some ways it's like asking the question how long does it take to learn how to use Excel? As Sarah pointed out, for basic things it can be quite efficient and quite quick. I've been investing some of my time lately in learning about calculated fields and how to do more advanced things with a very limited data set. So where in the past I might have added to an Excel spreadsheet and made new columns that have different calculations in them. All of this can be done off of the very basic raw data. And that kind of calculation certainly takes, I would say for me, a couple months of getting familiar with Tableau for other needs before I thought, okay, now I think I understand this well enough to go into the more advanced features. But I do think that for just getting a sense of how the software works and as long as you are working with a data set that you are already familiar with so there are no surprises in the data set itself and you understand that. If you have a limited objective of just creating a graph without interactive features and publishing that, that it's certainly doable within a couple of focused one-hour sessions with the software. I would agree with Sarah that the online help is really great at getting you started. But it really is the more you put into it, the more you'll get. Thank you. Rachel? I have a similar reaction. It's pretty easy to get started with some straightforward views. I'm a fan of the bar charts. People have seen my little blue bars in many of my visualizations. They're pretty easy to do. They're pretty straightforward. So getting started with some straightforward data is really rewarding because you can do that pretty quickly. I also started in 2012, I think, in the fall and I went to a two-day training that turned out to be really useful. It was two full days and it really got me immersed in a whole set of things that I wasn't using immediately but had a little bit of context for how to go and get it as I had projects that called for those kinds of features. So whether it was filtering or customizing marks or how to label doing some different things like that that made it a lot easier, I think, having had that initial two-day training. I do look up regularly. I Google my answers, you know, what I'm trying to do something. There is great help online. And I'm also been really grateful for their online support. So part of your $300 annual fee for your desktop professional comes with the ability to put in tickets and get help. And they will call you and talk to you about your ticket, which I just love. So not only is there sort of a very quick turnaround time in response to any email that you put in but that you can send data, they can work with things, they'll call and follow up if need be. So I found that really useful when we were trying to do some of the calculations that others mentioned that I couldn't get them to work the way I was expecting and there was really good support for that. So yes, I'm still learning. There are still things that I can't do and I'm excited about learning more. So it's that mix between what you can get up and start doing right away, which is satisfying. But then there is more to learn. Yeah. And just to give a little bit more diversity to this, we've been historically using SPSS here in the aerial offices and of course IBM SPSS now has a predictive analytics environment so I ended up going to a half-day workshop to get a sense of what it does. So I'm finding, yes, with a few hours you get the sense of what you can do but you do need to spend the time to work with your data. These environments can be relatively simple to do something and get something quickly but as you incorporate more data, you want to relate more sources, it becomes more complex. Which brings us sort of to the third question, the third item we're going to discuss here before we go to the detailed ODBC demo that Sara did for us. People wanted to find out a little bit more about the interactivity you can have with Tableau. Does it do interactive graphs? At what level can you manipulate the data and how do you deliver the data? So yes, yes, yes and yes. That is my answer at Ohio State. And I think if I understand this correctly, I think that what we typically view as interactive graphs in the Tableau environment is being able to filter your data with your graphs. So Tableau has a function where you can set up a graph and a bar chart and the default would be all of your data. You can add a filter for a particular field and when that filter is on, that allows you to select different data elements out of that filter. I don't know if I'm explaining this right. So like say if U.S. State was one of your filters, you could say I only want to look at the data for New York and Ohio and everything will automatically recalculate on all of your graphs with just the data for those two states if that's what you're interested in. And so the limit and what you can do with that, there's also some features called action filters. This is the thing that I've been having a lot of fun with lately. And what that is is you can create a dashboard and then you can add these action filters which say when you click on this particular graph, it's going to, or if I'm going to click on this one particular filter, it's going to change the data for all of the graphs on my dashboard. Or if I click on this particular graph, it's going to open another graph on another page. You can click on a particular graph to open another window with a table with more data. You can click on something with an action filter that takes you to a website. You can embed a website. So I've seen a dance faculty member here at Ohio State who had this nice little interactive graph and when you clicked on certain touring troops, it would open up a YouTube video within the dashboard window with a dance for that particular touring group. So there's a lot of different things you can do to have interactivity with a Tableau dashboard and a good person to look at if you just Google Ryan Sleeper, S-L-E-E-P-E-R, I think. This is a gentleman who does a lot of work on Tableau public with sports data and he has an incredible number of creative ways to package and share interactive data that's really engaging and compelling and tells a really good story. Wonderful. Jeremy and Rachel, any comments on the interactivity elements? I'd start by just agreeing with the yes, yes, yes statement from Sarah. I think it's the interactivity that makes Tableau as exciting and as viable as it is for use here at UBC Library. I want to qualify what I said before when we start talking about adding interactivity, certainly that very limited amount of training that I described to get started. It takes more time to build that interactivity into it. But as Sarah pointed out, the filters are a relatively easy way to get started with interactivity and we've used that here at UBC Library to create kind of generic reports. So for example, for circulation that have a lot of underlying data and we don't necessarily know how the end user is going to want to look at this but we give them options to filter by item type and by location and by branch and by checkout range and by shelving location, whatever the options may be. And essentially what we're doing is creating a space where they can kind of build their own reports using these interactive filters. And so with that method it's not necessary for all of the end users to learn how to use Tableau. We just take advantage of the skills of a couple of publishers, a couple of people working with Tableau Desktop Professional to build in some of that flexibility so that a single report potentially answers a fairly wide range of questions. So I think the filters are definitely a powerful feature. There is one other aspect of interactivity that I've been exploring recently and that's by including parameters that the user, when they're looking at the dashboard online, they can type in their own parameters that then have an effect on the calculations in the data that they see. I'll give you a quick example here. Here's an example also of where you don't necessarily want to make your data public. But I've recently put together a dashboard to help with the allocation of our budget for student employees. And so we have a spreadsheet that has the budget request, the amount requested from each of our units. That's subdivided into various requests. So we're looking at probably 50 entries altogether. And on the dashboard you see a list of these. And then there are several parameters. One is the student budget itself in a given year. So you can enter that. And another parameter is the percentage of the subsidy that we may get from the university because there's a program that subsidizes some of the student wages. So you can type in a percentage there, for example. And in the example I used I typed in a student budget amount, and then I just plugged in 70%. And what it does is it calculates how far we would be over or under budget if all those requests were accepted and we got 70% of our request for the subsidized funding. But you could look at that and say historically, well we've actually gotten closer to 85% of the funding. And the user can type that in and have that calculate and essentially change all of the underlying calculations to show how far you would be above or below at 85%. So it can become a bit of a what-if planning tool and can help with some calculations that I think would be pretty complex to try to do in a spreadsheet. And in this way we can distribute it to a working group and have members of that working group plug in their different scenarios and see what happens. Wonderful. Rachel, any comments on this aspect? One aspect of interactivity that we use the most is working side by side at my desk with desktop professionals. So there are a number of filters and interactive elements that we use for data we've published to public like library or user group or loan type, different things like that that people can check and use those filters very easily themselves. But when they want to ask other questions they can't do it unless all of the interactivity is anticipated and built into the view they can only do what's available and then what they have access to through that public interface. So there's even the asking questions and interacting with the data will often happen at the desktop level, working one-on-one. But we also, this may sound so plain, but I'll take a copy, I'll take either a snapshot or I'll take the image and I'll put it in an email or I'll send it to somebody or we'll put it up at a meeting without the interactivity. And in part it's because it's difficult to get the kind of interactivity in a presentation setting or in a room where you're trying to convey more limited information and the limits of trying to share the information. So I think it does have great interactivity but some of the deepest interactivity is happening at the desktop level. It seems like each one of you has brought to the table all different elements of how you can work interactively with Tableau. This is wonderful. Excellent. Thank you. So it did come up at one of the presentations. People wanted to find out a little bit more how the ODBC connection works in Tableau and Sara Murphy has created a bit of a walkthrough. Sara, can you take us? Okay, let me get situated here. So Martha asked me to talk a little bit about how we connect to Iliad using an ODBC connection. And this was in relation to some slides I shared at the first presentation or first webinar. And here's one of them. We took a look at our Iliad borrowing and at Ohio State our Iliad departments do not align directly with our university's academic departments. And our librarians are not necessarily, we might have like a biological sciences librarian but that individual is assigned to seven different departments that are related to biological sciences. So we put together this visualization using a number of different sources that would allow us to filter our borrowing requests based on our users' departments in a way that was meaningful to us. So we'll say history, and these are basically all of the history departments at OSU that are assigned to this one librarian. So this is just another part of that visualization. So again, to create some of these visualizations, we're blending data from a number of different sources together to make them work and make sense and to create a richer visualization that's meaningful to our user population, which is our librarians. So anyway, so if you're within a Tableau desktop professional workbook, this is the page that you would see. And I mentioned a little bit earlier, it does get easier the more that you use this tool. You start to learn the cadence of Tableau and how these pages are set up. So for instance, right here we have a data card, our dimensions card. This is where all of our categorical data will appear once I connect to the data. And our measures card, let me get the arrow right here, which is where all of the measurement data is going to show, the money, the usage data, anything that's a number is going to show there. Some of these right here, this is where you put up, this is your filter shelf. You can drag and drop dimensions and measures under your filter shelf. The marks card is where you change how, the appearance of the data that you've dragged and dropped onto your dashboard and your worksheet visualization right here. But anyway, to connect to the data, I would just up here click on connect to data. And then this window will open. And I would scroll to the bottom because I'm looking for the ODBC connection. It's always the very last one. But up here it shows that these are all these different kinds of data sources you can connect to directly using Tableau and they add more every year. Once I click on that, I select from this DSN here, I select what I've named OSU Link. This just opens up my ODBC connection box that's on my desktop PC. I have a couple other systems I connect to, but I connect to this and I say OK. It's going to open another window where you type in your login information just like you would normally if you were using access. And then I get to this window and I have to go through and find the OSU database, OSU data for us, other institutions, I'm sure you know your own name, type in the database schema, and then start finding the tables that I'd want. So I always start with the transactions table and then I move on and I connect the transactions table to the users table. And right here we have a basic Venn diagram with a inner join. I need to change that though to a left join and that means that all of the data in the transactions table will appear and only the data that's present in the users table will join to that transactions table. I have to find a unique data element that's going to link these two data sets together. Let's see. Next, because this is such a large database I have to add a filter or it would just crash on me. I would never get any data back. So I click here on the add the filter up here on the right. And then I keep clicking right here on add and I usually say NVTGC keeps OSU. The process type I just want to keep borrowing. I want to find all of the transactions that are marked request finished. And then this is the really important one. I have to limit my date range for what I'm pulling. So the last time I went into this database I only looked for one year. And then I say okay. And the last thing I think I need to do is I want to change the connection from live to an extract. I tend to mainly work with extracted data. You can refresh and extract a data extract at any time within the Tableau workbook. But this is going to improve my performance if I'm working on an extract rather than on a live connection. So I do that. And it's going to ask me to save that extract to my PC and I'll say yes. And then I'll hit update now. And then I'm going to hit go to worksheet. And all of these fields right here will populate with all of the data. And when I get to the worksheet it looks like this is what I mean by I have to save my data source, my data extract. And then once I get to this page I might have all of my data for the transactions table which is right here. And the users table right there. If I click on this arrow right here it's going to open up and it's going to show me all of the fields that are within that table. Down here I have my measures. And then I just start dragging and dropping fields onto my shelf for the columns and rows and I make the visualization that you saw. A couple of slides back. So that is basically how you make an ODPC connection in Tableau. Excellent. So we have a walk-through for those who want to try that. Thank you. And any more questions anyone may have? We have a few more minutes and feel free to click on the chat box. Also I would like to invite all of you to join the ARLSS group and share their different visualizations you have used over the years. It doesn't have to be done only with Tableau. I think we've had a couple of postings of people using other interesting software pieces they found on the Internet. Clearly Tableau is more of an enterprise solution but it's not the only visualization software and I think the more we can share different types of visualization, the more adept we will become as a community to sort of figuring out new ways of exploring relationships and understanding really more in-depth the very rich data we have in place. I do have sort of a closing type of question. We have talked a lot about organizational data. All of you being assessment librarians, you are tasked to use data to support organizational decision-making. One of the early questions asked with a notion of offering it as a service to students but even beyond that, the notion of doing the data management and creating the data specialist specialization in our libraries, do you see this kind of software being utilized with some of the data management service development you are doing in your institutions? And if you don't see it, why not? Is it just a matter of time? I would say yes but I'm not sure that I can speak to how that would happen but I would think that it would be really great for that. And I know a lot of the data management roles are still being defined in our institutions so there is an evolution there. Jeremy and Sara? Yeah, I have talked a bit with some of the data management folks here about whether there is a role for it. And I think the way our services are structured right now, the kind of benefit that I think Tableau potentially could bring to that data management is helping researchers work more effectively with taking advantage of the visual element to understand their data. And I think at the moment the library is not prepared to really – unless a researcher has that knowledge I don't think the library is equipped right now to be able to teach the Tableau skills to make that happen. So I don't see it right away but that's not a – I guess I would agree with Rachel that I'm not sure exactly how but if the library sees its role as helping not only to manage and format the data and provide metadata but also to explore the data then there might be a role in helping develop faculty expertise in the use of tools like this. And Sara, would you like to comment on this aspect? I know it's sort of – and the data management services is a new area for our institution so it's evolving. Yeah, I've – as I mentioned earlier I've worked with our data management librarian to do a data visualization workshop for our library staff. Honestly with – I would love to do workshops for our students but with all of the other things that we're tasked to do right now I just – I can't make it happen. So but I think it was – my personal opinion is I think teaching tools or partnering with another unit on campus to teach tools such as Tableau would be a great opportunity for libraries but again you have to have the manpower to do it and the right kind of space and tools and hardware. I'm not sure that we're there yet at Ohio State either and again data visualization means very different things to different people in different constituencies. So I'm a, you know, a Tableau zealot and I openly admit it but it might not be the right tool for someone in another discipline, academic discipline. Yeah, different disciplines have their own norms. Yes, Jeremy? So if I may add one thing and I actually thought of this earlier being aware of how visualization tools like Tableau work with the data and how it needs to be formatted in order to work well with tools like that I think is really important on the research data management side and I think probably the kind of formatting and recommendations that a research data management program is making sets people up quite well to be able to apply the visualization tools but I think being aware of that interface are the things that we can do to prepare the data in such a way that if people want to use visualization tools with it, it's well prepared that I think could go a long way. And I think if anyone's interested in following up a bit on what I mean with that looking for a Google search for things like how to shape your data for Tableau or how to prepare your data for Tableau will give you an idea of things you can do to make the visualization much more intuitive and smooth the process. Right, that's really important but it's also helpful to remember too that one of the beauties of Tableau is that because it allows you to blend data and to do things like we haven't even mentioned about assigning aliases for your data and creating groups for data, so for instance if you have a really messy data set that one time it will say Thompson Library, the full name, and the other time it will just say I don't know like Tomp rather than the full Thompson in abbreviation. You can actually use features in Tableau such as the aliases and the editing and creating a group so that you can group data of that client together and just assign it all a new name rather than spending hours cleaning Excel spreadsheets with data. Another thing that I do with data that exports really messily to CSV format and then when you put it in an Excel spreadsheet it gets the columns all messed up and you spend hours fixing all the columns, etc. I now rather than deal with that I just export maybe one or two fields at a time and the one field will be a common unique identifier and I'll just join all that data back up in Tableau and I've just saved myself all of that time from having to clean that data and I would love to learn more about data cleaning so if anyone out there has a webinar or something to ideas on data cleaning and fast automatic ways to clean your data that would be great but I used Tableau right now to avoid having to clean some data because of the way I can use the tool and I find that really, really helpful. The metaphor I want to close with is a picture from the Minority Report movie where there is the actor there working on a transparent glass surface that uses his arms to pull different sources of data to combine them to create joints and this is more of a futuristic science fiction type of environment but in some ways this is the kind of combining and massaging and cleaning of the data that we are doing these days with some of the things you brought to the discussion today. So I really would like to thank all of you Rachel, Sarah and Jeremy. It's been wonderful to have your work captured through these webcasts and these webcasts will be available on the ARL YouTube channel and we will continue this conversation. Visualization is here to rule for the future. Thank you. Thank you. Thanks Martha.