 Well, hello, everyone, and welcome to the visualization and Tableau workshop conducted this afternoon by Kevin Fomalant. I am Sue Boffman from the Association of Research Libraries, and very pleased to welcome everyone this afternoon. It's great that you can join us today. If you have attended one of these workshops, you know I have a little bit of a spiel. And by the end of our series, you'll probably be able to recite it with me, but just bear with me while I share a couple things that I would like everyone to know about. The Research Library Impact Framework Initiative has been underway for over a year, and we have a variety of teams who've been very busy exploring a series of five questions. And these questions relate to space, diversity, equity, and inclusion, special collections, and researcher productivity. One of our overarching goals for this initiative is to help us understand how to address the most pressing questions that libraries have with regards to value and impact. And this initiative is being funded by an IMLS grant, and we're very appreciative of that. As part of this initiative, our two consultants, both are with us today. Kevin Fomalant, who will be leading this workshop, and Margaret Roller, have developed a series of workshops on qualitative and quantitative research methods. And our goal for this workshop series is to help library staff develop their skills and expertise in conducting research at your libraries. So today's workshop is a part of this series. Today's workshop and the same presentation on Thursday will be recorded, and we will share the recordings and Kevin slides and other documentation with everyone. And you are very welcome to share this information with any of your colleagues if you would like to do so. So with that, Kevin, I'm going to turn the virtual podium over to you. Thank you, Kevin, for being with us. Thank you, Sue. Thanks for the introduction. So let me just share my screen. So welcome to Tableau for visualizing survey results. Thanks for taking some time to learn about this software that I enjoy working with and that I hope that you will too if you find the chance that you, if you find the chance to at your home institutions. And so before I dive into some of the technical details about Tableau, sort of this presentation, I have fewer PowerPoint slides than usual. Since I'm, you know, sometimes people haven't been exposed to Tableau very much so I wanted to give you an opportunity to see what Tableau dashboards look like. So that, you know, if you go online you'll be able to recognize them when you see them and see what you like about them or what you think you might be able to do better in your own work. So I'll be doing a demo, both of the more detailed non-survey related Tableau dashboard and then one workflow related demo where we're taking survey results directly from the source and then putting that data into Tableau and making the visualization. So you'll get the best of both worlds for that. So I got into Tableau when I started a position doing survey research several years ago. And by that point, Tableau is already a pretty mature software. It came out about 15 years ago now. It was really developed in response to a need in the market to allow people and companies to create visualizations that without having too much coding to really any at all. And to, you know, be, grant them more flexibility than you would have in Excel. We all love, know and love Excel. You do amazing things with Excel. But it's also, but it's fundamentally a spread, it's a spreadsheet, whereas Tableau is a visualization platform or you're building on data that can be stored in spreadsheets. And that, you know, I work with, I work for an organization that does large survey projects. So I'm used to creating dashboards based on large amounts of data and then pushing those to a server where users will log on and actually see the, see the dashboards online. You may have that set up at your home institutions, but it's just as widely common for individuals to use Tableau desktop, which is what I'll be working in today. Where you're creating the visualizations, you're creating the dashboards on your home computer connecting to the data source, and then you're either using it for yourself or you're sharing it with a small group of colleagues. There's lots of ways that you can use Tableau. And that's one of the advantages is that you're actually, you're working with software that's developed for individuals to show in a small group but also that's been scaled up to the larger corporate level. And so feel free to interrupt me with questions. It would be great if you could do voice questions I will monitor the chat at the same time though. So please feel free. Tableau, no coding needed is probably the biggest example of why people like Tableau. So you can you can go from Excel, create a dashboard without any need to use our Python necessarily to format your data. I'll show you how you can do that today. So once in Tableau that you can create out of the data that you've imported, you don't necessarily need to have all the fields correctly formatted once in the spreadsheet before importing. And a lot of the language in Excel will seem familiar in Tableau. I like that Tableau, you know, you can connect to a spreadsheet and connect to relational databases for survey platforms it's probably just a, it's probably just an Excel or CSV export, but there are definitely lots of times when I'm working directly with a with a SQL backend. One underrated aspect of Tableau is that you can combine your visualizations into one dashboard. And when I'm writing, when I'm generating a written report in a Word document, you know, the visualizations are separated out you have to scroll through it and sometimes it's hard to hold in your mind what you've just seen. So the advantage of Tableau instead of interspersing those visualizations with text is that you're combining them all into one dashboard so that you can see the information and compare it against itself. Anyway, Tableau is actually a report replacement, depending on the forum that you're working in. And one thing that's not well known I think about Tableau is that you can share it with your, with your coworkers with your colleagues with who don't have a subscription. Of course, it is a subscription service you do need one to work in Tableau desktop but there is a way to share what you develop freely with other other people. And I'll show you that at the end as well. So today I'm going to go through what Tableau can do in Tableau desktop. I'll do a demo. I'll talk about data structures which for me is a very important thing in Tableau. A lot of times when people are, you know, are frustrated with Tableau or they're frustrated with any visualization program it tends to do with data structures. Since generally the rate limiting step is putting the data into the right format so that the visualization can be done. And creating the visualization is really the fun part where you're moving graphics around and you finally get to see the insights into your data that you've wanted to see from the beginning. So formatting correctly, putting in the right structure is the rate limiting step. I'll show you how to import data into Tableau which is very simple. I'll do a demo with survey data. So you can see a simple workflow to put together a dashboard to present your survey results. That dashboard will include multiple visualizations, which I think is important and then a way to share Tableau dashboards with your coworkers. What can Tableau do? So when you're working with Tableau, there's a few ways you can go about it. And one of them is to actually hold the data in Tableau, but it's actually not the case. The Tableau actually has to store the data itself. It can draw on the data from the spreadsheet from outside or from a relational database without actually holding the data in it. It's important for a number of reasons. For data security reasons, you want to keep the data out of Tableau. You can do that. On the other hand, if you want to share the data and you have no concerns about that with colleagues, you can keep it in Tableau. And then if there's a live connection, that is, if you allow Tableau to always be connected to your data source, it will update as you change the spreadsheet. Although with some a few steps along the way that depend actually on the platform itself more than Tableau. So Tableau can process and transform data. The workflow I'll show you that are similar to a lot of data analysis workflows is that you'll be doing transformations in both the spreadsheet and in Tableau itself in a reproducible way. So you can save time and not get frustrated on your path to creating these visualizations. You can create new calculated fields in an intuitive way, visualizations. You can create multiple dashboards. There are three components to Tableau's visualization process, creating a sheet, which is the initial creation of visualization, putting those sheets together into a dashboard. And then finally you can create a storyboard from those visualizations and dashboards if you choose. I find that that's beyond the scope of what I usually need since it's more of a, it's what you would use in a presentation compared to what you would use in a dashboard. And then one thing that I like that's important is that you can filter information by group and permission level. So if any of you are interested, I can show you a way to filter dashboards so you can limit the people who see the individual visualizations. That's a little bit beyond the scope of the demo that I'll show you, but I can help you with that if that's something that your institution is interested in. And also, my favorite part of Tableau is the filter so that you can view, you can really drill down into the data that you're presenting by selecting a subgroup from a, from a filter. That could be a demographic group, an age group, or just down to the, down to any granular level that you want to see detailed information about. A few of the standout features that I like I mentioned filters already. When you are in Tableau. And this is something I actually think is not matched well in other software that I like and use all the time like are is that geocoding is automatically done excellently in Tableau. Maps are generally hard to create on your own. I've tried it myself I've tried stealing some Google Maps and and overlaying information on top of that. I've used some some map specific commercial software that requires a lot of programming ability. What Tableau does it actually has, it's really done everything for you all you have to do is type in the right state names, the right zip codes, and it geocodes it for you and you don't have to create the map itself. That's one of the major advantages that I like. There are many ways that you can represent information in Tableau and that's through, through marks, which I'll show you in the dashboard itself. So in the, the map that you see on the screen right now. It's, I think it's a population map so the larger the circle, the larger the population that country, and then at the same time the color I think is birth rate. There's only two dimensions that are easily represented in that visualization, or as an Excel this would be a challenging visualization to create not just with the maps but with the multiple dimensionality. And some of the drawbacks that I do want to talk about no software is perfect. So when you're using it you need to be aware of some some things that will trip you up, or when you're thinking about sharing data with colleagues, what are some potential issues that you want to share with colleagues before you go ahead and do so. So I did mention that you can connect directly to your data source or spreadsheet. There is a way to refresh it, but it does. It, it can be delayed. So many times you'll actually need to close out of Tableau and come back into it to refresh the data. And again that's more of the sort of the lack of communication between proprietary software. But in the end, it is something that updates automatically in the sense that you don't have to redo the visualization kind of like an Excel. You do need to do transformation, particularly for survey data, I think even more than other data types you do need to do some more data processing. Because of the data hierarchy that is innate to survey data. I'll show you how to do that. I'm going to show you data sources. And when I'm creating templates for larger clients, I, I want to make it as general as possible, so that I can reuse that that template for, for a future client. And in that case I could just change the data source and connect that to the, the template itself. I believe that should work without any hitch, but I will say that, you know, changing the data source does tend to cause some bugs because of the data structures in your spreadsheet any small difference, sometimes even unseen differences that you can't tell from the spreadsheet itself will cause an issue with the visualization on Tableau side. What you should be aware of if you're planning yearly, yearly work is that you can create a template but also take some spare some time to, you know, prepare for the next year, the next year's report the next year's dashboard, so that when you hit those issues, you're prepared for them. And one thing that I don't think too much about but I guess it can be an issue if you're sending your, your workbooks to other people. Tableau workbooks are very large it is large software. Any visualization software is going to take up a lot of space. So when you're downloading Tableau desktop or Tableau reader, excuse me, onto your desktop be aware that it will take up a decent amount of space. And then it's difficult to use with other large programs at the same times. So it's best to focus on Tableau itself. Do your work and then move on to the next step in the process. Tableau desktop is the subscription software that I'll be working with right now. It includes all the fundamental features that you use in Tableau. And then I alluded to this earlier there are two types of workbook files that you can use in Tableau. So the twib, the twib X files, they're called. And so the twib file doesn't include the data with the, with the workbook itself but the twib X file does. And this is important that if so that if you want to share data with your colleagues you save it as a twib X file, the data will be embedded into that workbook. So you send it and they'll see that data itself. If you save the if you save the workbook as a twib file, it won't have the data with it so you're just sharing really an empty visualization if you send it to a colleague. So you can share dashboards through Tableau reader or the newer versions actually Tableau viewer. I'll show you how to download that so that you can send the visualization that you created in Tableau desktop to your colleague, ask them to download Tableau reader, and they can, they can view it without being able to edit it but they can see it. And then, you know, once I share the dashboard that I created for this, this workshop, you can use Tableau reader to download that that workbook and play with it yourself. So I commonly actually use in my work, Tableau server and Tableau online, which allow you to post your workbooks to a web portal and then do some more work with permission structure. So this is the, this is for wide, wide use so you can allow, you know, dozens or hundreds of users to log in at the same time and view your dashboards, which is something that Tableau desktop doesn't have. Only one person can use the workbook at the same time. I'm going to switch over to the, this is a demo workbook. I want you to see what Tableau looks like. That's not what it can do really. That's what it looks like in general. So this is what happens when you, you open a workbook you start, you add in a connection. So this is the world indicators connection file. You add your connection here so you can connect it to a Microsoft Excel file, which I imagine is what most of you will be doing will be doing and then also some relational database connections at the same time. And so this is the, the data that you've imported by column. So you can see it handles a lot of different types of data. Well, actually, as you can see here, it's in my other workbook. I think it's, this is an older version. There's a nice data interpreter that that I noticed the last time I used Excel, or sorry, the last time I used Tableau actually did a nice job of fixing Excel data without me doing a lot of processing. So I'll show that and show you that in the next one. And then this is the live connection. So this is the one, the connection that connects directly to your data so that when you make a change in Excel in your back end you can hit the refresh, and then this will update what you see here in Tableau. And so this will show you click here and it will show you the data type. Common ones numbered date string and then also geographic role which actually is quite sophisticated and go down to the congressional district level. For, for many of you I imagine it's state or zip code or even country that you might be most interested in. And then if you, so you actually have to dump the table into the, into this visualization here. So this is the data table itself from the, from the world indicator source. But a lot of times when I'm working in Tableau I'm actually joining a few pieces of data together. And if there's a so if you say you have two different sources of data. You have economic information about different countries say you have other information about the country itself. Other demographic characteristics in a different spreadsheet, which you can do is create another source to another Microsoft Excel spreadsheet. And then as long as there's two columns that have country information, have country information you can join that data on that column. And you don't even need to be named the same way, something some other software doesn't do so good for Tableau for getting that right. And that is, that's useful. You know, if you since you don't necessarily want to have to do the combination yourself in an Excel. I don't have to copy and paste into another spreadsheet you can just dump both both spreadsheets into into Tableau and join them. So that's the data spreadsheet you'll see when you open up Tableau. So the color at the top, it's a little bit subtle I think you can see it though blue matches the connection. But you'll see that there's also some fields here that don't have any color on them at all. And that's because they're calculated fields. So what you can do if you want to get one here. So if you want to calculate something about birth rate. Maybe you want to say that there's a correction for birth rate. Your numbers are off by 1% and you can add that into the calculation and then name it something birth rate correction. And now you have a calculated field. So you can use you can use both fields in the visualization itself. You can also toggle to the sheet. So you're looking at now is a Tableau sheet, which has, which is really where you're going to be doing most of your work and creating the visualization itself. So on the left is the data tab, which will show you all of the pieces of information that were in your data source tab. So green the green color means that it's a continuous, and then blue means it's categorical. So that means, and then, importantly, the marks this is where you'll be doing a lot of your work the marks portion of Tableau is the type sort of the way that you'll be visualizing the data itself. So there's color size label detail and tool tip. And then you can choose whether to make it a bar graph or here what we've what it's done is to make it a map. And then you can choose either types of visualizations which you can see I'll show you a few more later. So like I think I showed you in the PowerPoint there's actually two dimensions that appear on this visualization. One is the birth rate by color, and then population by size. And then it comes to these two measures they're called in Tableau. If something is numeric, and you use it to calculate a statistic it's a measure, whereas a dimension is more of what you would think of as something on the the x axis. So it's a little bit, it's a bit more complicated than that, but I think is measures as the y axis values and dimensions as the x axis values. And because you can actually switch them. Like if you want to treat. If you want to treat population 65 plus as a measure, you can do that here in Tableau and that will change, and that will change the type the data type that's native to Tableau. So the marks. You can tell what color size and label are the detail is useful. It only shows up if you hover over the item. But it is nice. I think to not have to hard code text into the map itself, and just use the detail to reveal the information that you want to see if they're more interested, one of the nice features. That's just filter got away from me. But you can add a filter, let's be a good filter. Customize the filter. And so, I can show you a little bit more about how this works in the survey spreadsheet. But one nice thing is that you can have the filter apply to the to one worksheets or all of the worksheets as the case may be. So when you're combining visualizations from multiple worksheets into one dashboard you may want it to be the case that the filter works for all of the visualizations generated separately, so that it's not confusing. So you can you can if you want you can have each, each filter applies separately to each of your visualizations in the dashboard. Just show you the technology one. So there's also an analytics bar, which automatically calculates some statistics for you without having to do much work. So the trend line is one that's useful there's already one on there. And for the fields themselves. You can actually. So, Tableau calculates simple calculations for you without having to do a lot of the hard encoding that you would have to an Excel or really just typing it, but you can, you can select your percentile if you want, instead of average, it's an automatic change. So there's some few simple, simple calculations that you can do just from a drop down menu that you would otherwise have to do in the, in the shelf where you would actually type out the formula itself. So, so just to be clear, these are the these are the sheets here and then the dashboards are here where you're combining visualizations. So now we know, you know, what what kind of what kind of dashboard look like, what is Tableau some, what are some Tableau functions that we can use to make our visualizations. Now I'd say, there's a lot of options I showed you there to use. Probably when you're getting started you might use 10 or 20% of them, you may not even need to calculate a field. Maybe you just want to create a bar chart, a map. Maybe you just even want to use a sheet and then put that into a dashboard and use that to share your results. That's totally fine. A lot of times I'll use Tableau for descriptive analysis, you know, before I get into a more in depth, you know, statistical analysis. So it's, it's good at the level of seeing your results immediately, so that you can, you get an overall sense of what you what you've gotten from your survey. But then also, you can come back to it later and you have more detailed information when you want to do a deeper analysis. Excuse me. And so one thing that I, I'm sure you've heard me talk about if you've attended a workshop that I've hosted before isn't that data structure. That is really what determines whether your, your project is going to be successful, whether your visualization will come out the way that you want it to. So it's something that I do focus a lot about because it's something that people get frustrated with, and we'll just sort of, they'll drop the project. I've worked with some students who they want to, they would just want to go back to Excel or something they're familiar with, which is understandable. But at this, you know, when you're working with something new, there's a few steps that you need to think about when you're creating visualizations and Tableau that will really get you running quickly into making visualizations where otherwise you might spend a few hours sort of struggling or searching online about how to make this work. And so specifically for, for survey data. You're sort of actually dealing with a two level hierarchy, according to Excel. Right, you have a bunch of questions, but questions themselves are actually a sub hierarchy within question, a question as a type of data. So you have question, you have a question questions, and then you have the responses to those questions. So because of the two tier hierarchy, you actually need to represent the data in a different way. In a spreadsheet, you're probably more used to thinking about data in wide format, where there's where the, the different variables are along the top, and then the values corresponding to those variables are below. Whereas Tableau actually prefers survey data in long format, where so the X, where so the X value in this would be actually be the question number. And then the value would be the response to that question. And so if you have a participant name or participant ID for each person in the survey, they would actually be repeating multiple times in the in the Tableau data source and that would actually be the correct format. Now it looks, if you're not used to dealing with that kind of data, it actually doesn't look right. You're thinking, why is that, why is that name repeated. That will, that will cause problems on the other end. And it's really in this way actually that the Tableau is more like some other programming languages, like our, for example, when you do visualization work in our, it also prefers long format. And so, even if it doesn't look intuitively right, it is important that you just, that you go with it and give Tableau long format, so that you can create your visualization. And then we can actually deal with the problem of repetitive participant participant IDs in a pretty easy way, once we create our visualizations themselves. So I just want to check in if there's any kind of questions as of now. Feel free to hop in. As you saw in the Tableau dashboard itself, what Tableau, what the data source looks like in Tableau itself. It does have an Excel like language on the right is actually an example of a level of detail expressions near taking fields that you've already input into your data source and doing a calculation that isn't one of the more common ones I showed you like some are average or distinct. I showed you that they're data types that are similar to to other ones that you've seen in other in other venues string numeric data geographic being the one that's really useful. And then calculated fields are those fields that are created outside of your spreadsheet. Yeah, I showed you to add a connection through spreadsheets. You know, at the same time, when you're adding a connection, you might find that when maybe I'm not looking at my spreadsheet in as much detail as I should be. I use Tableau to troubleshoot errors in my data, whether there's missing values that I didn't see. Or even I'll use a visualization, a sheet visualization itself to try and find problems with my data. Maybe there are too many mills. Maybe there's a misspelling, or there's actually an extraneous value in the response data. You can rely versus extract. Extract is the one that takes the data from your data source and puts it in Excel. So that once you hit that extract button, it won't update until you put it back to live. And then for importing into Tableau, there is sort of a choice that you can process the data before or after creating the connection so you can try to rely on Tableau to do all the work for you. So for survey results, I recommend that you do some processing, some pre-processing in Excel, and then do the data pivot in Tableau itself. So for preparing results for survey results for Tableau and Excel, I have a screenshot here of the survey results from what we did in the last workshop together. So any export you get from a survey platform is probably not going to be perfectly compatible with Tableau, so you need to do some pre-processing. Sometimes I do it in R if you want to have some program experience, but you can also just do it in Excel. I've also done that for large clients to format data in a way that would be useful for visualization. And so that includes things like paying attention to the question text, numbering the questions, knowing what sorts of response types you're getting, whether it's multiple choice. So if you think about it, the way that we, and I can show you the survey, the activity survey that we did last time as an example. So it's easy to think about a bar chart for a multiple choice question, right? So question one, you would just have a bar for each one of the response choices. But for a multi-select question, it's not as obvious. So the standard way to deal with this in survey research is actually to subset a question like question three into 3A to 3F. So that's something I actually did manually in the Excel spreadsheet. Since there are people, since it's more of a, did you have this experience versus you didn't? It's a type of question. Instead of a multi, a multiple choice. So when you, when you're representing that data, I think the more intuitive way to look at it is whether that person selected R or not, Tableau or not. Not necessarily piling it into one question, because then you would get, it would be hard to represent a person who chose, you know, Microsoft Suite products and Tableau and R. You want to know, you want to know what each, what people think at the group level about each individual product, not necessarily whether one person likes these three or those two. So the standard ways to divide those up. So that's our question for, so for 3F, or for 3E this would be, for writing questions, I actually typically don't represent them in Tableau. You can represent them indirectly if you want. I would suggest then coding that data into categories. Maybe for this question, what type of software MATLAB comes up a lot. So you would be able to, to code that or code any misspellings, something like that, into MATLAB in pre processing, and then fix that in the Excel spreadsheet and then push that to Tableau. So you're actually thinking about a few things. And this is why this is the complicated part of the process and not the fun, easy visualization part is you're thinking about survey structure, how to pivot the data so that it works for Tableau, and then thinking about how to divide up the questions in a way that takes it from what it's intuitive for a user to what's intuitive for someone, a researcher to look at. And what you're really doing is bridging that gap, which is one of your most important jobs. So, yeah, so I mentioned the transforming survey results. It's called pivoting. So in the example, I'll show you briefly you'll be changing, I'll be changing everything from long to wide format. So the first thing to know about survey data and is that there are, in addition to data related to responses to the various questions there's something we call administrative data. And administrative data is data that's really that describes the individual it's the individual himself or her herself, and not the answers to the questions so it could be there, their individual participant ID. It could be the time that they took the survey. It could be the institution that that they come from or department. Now you can get you can get administrative data from individual questions. But the important thing is to separate out administrative information from your mind in your mind from from question data since it's really only the question data itself that needs to be pivoted. So that you can represent your survey result by questions separately. So the other advantage with thinking about administrative data is that this is something I'll cover a little bit more in the third workshop in the series. This is cleaning and processing the survey data itself so for deduplication, you know, some in web survey platforms, a lot of times you will fill out the survey more than once. And in that case you want to delete the survey that has been less completed compared to compared to the the second, the second attempt at the survey. But it's important to you think it's important to think about administrative data. So if you see the same participant ID the same IP address if you're using survey monkey, you can use that to weed out extra extra responses extra survey completes. And so, for creating you know a dashboard specifically from survey results. I have a screenshot here of the result. So that gives you a sense of what of a simple, you know dashboard that you can create from a library services survey, the survey that I showed you the PDF of. So the steps I'll go through a processing the data and Excel, pivoting the data and Tableau, creating the visualizations in the sheets themselves and picking the data points in the marks that I want to use that merges visualizations into a dashboard and finally saving it to a spreadsheet. I just want to show you what, when I started on Tableau, what I tried to do. So I have the raw spreadsheet. I'm just going to, just going to put it into Tableau and see what happens. So I have some data. Okay, so this is the raw file. So Tableau automatically selects the first line in the spreadsheet as questions. And so that's just what it took from survey monkey, which I don't find very helpful. And at the same time it sort of mixes. Depending on the question type of it's a multiple choice question. So it'll put the question number, and then the responses. Sorry if it's a multi select right this is the multi select question from the PDF. And then it will just put, you know, whether the person selected. Yes or no to that. But it'll just assign it a letter and a number that just isn't very useful. Whereas the question itself is over here. I'm already sort of just by looking at it, you think, well, maybe I'm having, I'm having some data processing problems, but maybe I can do a visualization anyway. Let's give it a try. So it's not going to do much. You're not going to get much out of it. And that's because the data is unpivoted. Tableau doesn't really know what, what format you're giving it. So what it does is it just adds up everything together and gives you. So represented, which is this tiny line here. And so if you're new to Tableau's, this is, and this is what I did the first time I used it years ago was just push my data in and then see what I could get out of it. Unfortunately, it doesn't get you very far. So what I'll do now is show you what to do in the Excel spreadsheet. So this is a modified version of what I got out of Survey Monkey from the survey that we took last time. So what I did is I took out the survey text, the survey question text, and I replaced it with simple numbers. And then for the multi select questions, I divided them into sub parts. So for the software question, I divided that up into 3A through 3F and then use just a numeric order. There's another multi select here for question six in the survey, and I left the demographic question by itself. I did add a location column for this visualization that I didn't collect, but it's there. And so one other question I get a lot is what to do with this with the survey question text. I actually take it out of the Excel file since there actually isn't a way to make a direct connection from the Excel files survey text, survey question text, and visualize that in a way that's useful. The best way to visualize the survey question text, I think, is to hover over so that when a user hovers over the question itself, QL1 doesn't mean a lot to them, so they hover over and they see the question text. In order to have that happen, you have to make a calculated field. You can't actually use the spreadsheet itself. So this actually makes things easier. It's a cleaner way to do it. So all you're doing is deleting some of the extra rows here that SurveyMonkey gives you, manually entering in the question numbers, and then making sure they line up with the responses normally. And you can delete some of this data if you want. I actually kept it in. It's part of the visualization. So even in large survey projects, I actually do do a certain amount of manipulation in Excel before importing. So we're going to use that. I've used that to create a results workbook. So this is actually the pivoted version. I went ahead and you can see it's actually pivoted. See if I can unpivot it. Okay. So this is what you would see if I had just made the connection. You'll just see the questions come in. But again, what will happen is the same thing. You'll just get that funny visualization with that thin bar that's not very useful for you. So instead what you do is you highlight all of the question numbers in the data source, and then we'll just repivot it. And then because I already named it, it called it questions, but you can you can rename this if you want. You can call it whatever you want, but we'll leave it as questions. And so now the data is in the format that you want so that you can create bar charts and maps and whatever you want. But notice I left the administrative data, the time and the time start and the participant ID here without pivoting it and that is fine. Another thing I did was I created a calculated field. I didn't do it in the data source so I'll show you that in the sheet, but I did create a calculated field here and the data source called survey duration. So I was curious if there was an effect of participant by the time it took to complete the survey so what I did is I took the time and and the time start. I actually did backwards. And that way I have survey duration in my calculated field. I think there's, there's a little bit of a source problem with the data but the calculated field will tell you what the survey duration is for this and then you can compare that against participant ID or the questions that they answered in the survey itself. So I created a calculated field there in the original data source. This is a simple bar chart that I created. And so what I did here is put questions and responses both in the column side. So that you have that on your, your y axis so there's a question on the y axis and the responses are on the y axis which was a little bit different from what you how you would think about it in Tableau. So what you're actually doing is filtering out the questions column that you you pivoted into right, we took the list of questions pivoted it down. So now we can filter by question by doing by doing this. So show filter it's already shown and pick something else demographic question. What I'm going to say about filters that is, I think it's for survey questions, is that it's better to actually not allow people to multi select. So I'll show you what happens and they multi select. If you allow that, and you get sort of a visualization that you don't have a lot of control over. And at the same time I don't. There's differences opinion about this but I think it's worth it to separate out the questions in the in the menu itself. And then you can download the visualization, especially when you're combining them into a dashboard with other visualizations you sort of lose control of your visualization. So what I'll do is customize it, and then I'll get rid of the all value. And then let's put it back to the way, put it back to the way it was. And then I alluded to earlier a problem with the way that when we pivot the data remember what we're getting is multiple data points for each individual in the in the long format. So the participant ID will be repeated all the way through the data source for each answer to each question. And so, if you don't take that into account, and you're counting participant IDs or you want to get responses per participant ID, not really in this one but in the, let's see, in the map. You need to count, you need to count distinct. Otherwise, you will get a large inflated number of participants right because the, the, the location is administrative data it's just one piece of data associated with each individual. Whereas, if you are filtering by question, right you're only going to have one participant ID so there won't be that inflation of the number of responses by participant. So that's one of the tricky things with Tableau with survey results is you have to be thinking about what's administrative data, what's question data. But the solution is quite simple. You just change the count of participant IDs. And this is called the shell for you can actually just enter in the, the formula, you change it from count to count distinct. I don't know if you can see that just count D. That's why you're getting the distinct participant ID so there won't be that multiple in your, in your visualization. Now, if I get rid of the D you won't really see it in the visualization you're just going to see the, the number, but I can let's see. Let's get rid of it. So if you count it, it's going to come up with a funny number. But if you do the count distinct, it should, it should change it correctly. So you'll get the, you'll get the color scheme that you want and the difference in the accounts of the participant IDs. Since you've eliminated the duplication of multiples across the white, sorry, the long format. So where I wanted to plot survey duration against a question number. So here, I have the questions and the responses on the y axis, but I want to compare it against the measure, which is average survey duration. So what I did is I went in here and I chose a measure that was easy to, to calculate without me having to go into shelf. So now have the average survey duration. And then the question text should appear when I hover over. When I hover over this and it does, but that didn't come through automatically what I had to do was create a calculated field, quite manually. It's an if then statement where the question is one. So question one being one of the data points from the list of questions, and then I just entered in the question text itself. I pulled it here made it a detail mark, and that way it will show up when you hover over the question itself. So those are the three separate sheets that I created for this. I do want to show you location though. Let's remove it. So it title does a good job of recognizing location. The thing is you just have to actually make it do it. So, select geographic role. This is the state province. So it will record it will actually even collect correct spelling errors if you've typed out the state wrong and prompt you to fix that. And so this is the dashboard so the sheet if you want to create a new sheet is there. A new dashboard is he a new dashboard is here. And so what I just did here was literally drag the sheets onto onto the dashboard. So you actually don't want to do tile do you want to do floating, which just means that that's more, it's more easily manipulated. And it's not. If you do tiled it'll load up the entire dashboard. So you want to do floating that way I can actually change the size. Make it the way I want it to go. It's the same for the other two these are both floating. So that when you're, when you send this to your colleague though, you can actually hide the sheets. And then you'll just see the dashboards they won't even know how you created it it'll just look like this great dashboard you've made. And then I actually made this filter I believe to apply to all of them. So you notice that the results in the survey duration charts changed but the respond locations didn't, and that's because you know the question number doesn't determine where the respondents is from since it's administrative data. So if you don't want that to be the case you can go here, select which worksheets you wanted to apply to. I'd like to keep it for all of them. There's the storyboard, which I don't use too much but it is here if you are. You can add more text really and not create a dashboard without too much explanation or you can add titles and things to your dashboard which is is useful by itself. And then we have a few minutes not too much time. Not too much work to go through, but to share your tab of dashboards you save it as a Twibex, your data is embedded but there is the concern about data security so remember that if you have created a Twibex workbook. You save your data as that way that it, it's not encrypted when you send it. So, it is behind a wall. It's not out there for anyone to see it's not an obvious way to pull it out. Just be aware that there is that I don't send Tableau workbooks with personally identifiable information PII to two colleagues by email. Unless it's some anonymized something you can do. What you just need to do is have a send your your colleague, the Tableau workbook that you've created in Tableau desktop and have them download Tableau viewer. Okay, send up send you that link. So that's the link to Tableau viewer. You can use that to view the dashboard that I workbook that I created. Once that gets sent out. And then if you're creating on your own to send your colleagues. What's important in Tableau. The most important step is, is data preparation process pre processing the data in Excel once you get your export from SurveyMonkey or from Qualtrics to assign question numbers to get rid of question text and to think about how to handle select questions versus multiple choice questions, pivoting your survey data from wide to long format for for survey data and Tableau using calculated fields to add in question text or to make extra fields. In addition to what's what you have in your Excel file, and then creating dashboards with multiple visualizations so that you don't have to have your user toggle through each of your, your creations. So that's the most possible to share dashboards with colleagues and that last point is to be aware of data security and that Tableau Tableau workbooks aren't encrypted. Don't send them by email unless you've anonymized the data, or it's data that doesn't have PII. I'll leave time for questions. So we have a few minutes left. This is the first iteration of Tableau for visualizing survey results. So there's another session on Thursday, but coming up in May the next workshop has to do with quantitative analysis. And so that's when we'll be getting more into the nitty gritty of statistical analysis, seriously processing survey results, and then looking at cross tabs. And so that's when you're looking at the concordance between questions, how and then some groups analysis, how different demographic groups answer questions or the reliability of the questions and then significance testing particularly for year on year changes. So I think I did get a question. Appreciate it. So for, yeah, just feel free to ask any questions you have about Tableau. It's helpful for you to see sort of what Tableau can do and then exactly what to do when you have survey data and how to process that. So yeah, just jump in. Kevin, this is Sue. I have a question for you about this and it looks like as I've listened to you that, you know, with anything practice makes perfect. Do you have some advice if you're brand new to using Tableau? I mean, just what to do with it, just play with some data and see what works and what doesn't work. You have some best advice on how not to be afraid of Tableau and just start working with it. Yeah, thanks for the question. I think that I think the first step is creating the most simple visualization that you can not get bogged down in the data processing. So what I would do is just isolate the data from a single question to import that data into Tableau and then make a bar chart so that you see that you can do it. You can repeat what you've done in Excel and then build from there. And then I think one of the more, if you don't want to do a bar chart, I think one of the more rewarding ones is actually the geographic visualizations since you really only need individual IDs and then states or zip codes and then you can simply drag that, change it, drag that into a sheet, select the geographic role and it will automatically create a map out of nowhere without without you really doing much work at all. So I think the idea is to get an easy payoff before getting bogged down in data processing. Great, thank you. Does anyone else have any questions for Kevin? And if you do have a dashboard that you like help with or a data processing workflow that you want help putting together, feel free to reach out. I can help you one-on-one if you'd like. Kevin, we may have some colleagues on this session who haven't didn't attend your first one, so we can be sure when we send out the slides and the recording to share your email address so that colleagues can reach you. But I would second that please reach out to Kevin if you would like some assistance or just have any questions. I know he'll be happy to provide assistance. Absolutely. Kevin, since I'm not seeing any further questions, maybe we can give our colleagues a few minutes back in their day. But thank you very much for this workshop. I would encourage the colleagues on the call if you have colleagues that you think might enjoy this workshop. Please have them register or reach out to me and I'll make sure they get the registration information for Thursday's session, which will start at 1pm Eastern time. With that, Kevin, let me thank you again for doing this workshop. Every time I look, learn about to blow, I learn something new every time. So thank you for conducting this workshop. And thanks to everyone for being here this afternoon. I will bring it to a close then. Just thanks everyone for your attention.