Lesson 9 - Storytelling with Data Presentations

Welcome back! You spent the Module crafting a research question, finding and cleaning data, performing statistical analyses, and creating data visualizations. You should have quite the array of insights by this point, but what are you supposed to do with them, and how are you supposed to present them? Well, it’s a good thing you asked, because that’s exactly what you’re going to be covering in this Lesson!

In this Lesson, you’ll be focusing specifically on making a visual presentation you’ll be able to share with stakeholders, while in the next Lesson, you’ll add an oral presentation component to round things off.

Creating a visual presentation involves creating a data story in Tableau. This data story combines your data and analyses with your research narrative. Think of it as a form of storytelling. Storytelling for data analysts is extremely important as a means of reaching your audience and communicating your findings. You’ll be going one step further and actually building a story for your Cliqz project in Tableau. If you haven’t booked a call with your mentor in a while, now’s a great time! Your mentor likely has a lot of experience giving presentations in a professional setting and can share some good examples.

It’s time to tell your story!

 

1. Data Stories

Data stories, or data storytelling, refers to the blending of data insights with an overarching narrative. This concept has been around for many years. Data stories can take all sorts of forms. They can be used to inform, coerce, change minds, forecast, or even lead an audience to a known conclusion. People don’t like to change their minds—even about known facts. Sometimes leading them through a presentation and making them ask the questions themselves can be enlightening!

It’s important to note that propaganda is also a form of data storytelling. However, this version often hides or obscures the correct conclusion by leading the reader to a different inference. Make certain when you create data stories that you use them in an ethical manner!

You’ve likely experienced data stories yourself in the form of infographics. Infographics are visualizations that include text, graphics, and data. The information they present isn’t interactive; rather, it takes the form of a static graphic that can be easily shared in online and printed material.

The infographic in Figure 1, below, displays information about Canadian immigration using a mixture of graphics and text to guide viewers through the data and tell an overarching narrative. You can see that, by 2035, a large percentage of the Canadian population will be older and ready to retire. Because the fertility rate is too low to replace all these retirees, immigrants will become the main source of population growth in the country. The way the infographic combines text and images with the data makes it easier to consume. It creates a story—a story that emphasizes immigrants’ importance to Canada’s future workforce. This is a case of leading the audience to a conclusion through storytelling. In this example, the Canadian government wants to prepare its population for the eventual change in demographics. This is one way to do it and foster good will in the population:

Figure 1. The visually appealing use of images, color, and text makes this infographic easy to consume and understand. Source: The Conference Board of Canada

Tableau allows you to take infographics one step further. Rather than making static images like the infographic above, you can use it to create interactive data narratives. This type of visualization in Tableau is called a data storyboard. Storyboards are useful not only for reaching an audience and communicating with stakeholders—they also make great portfolio pieces to demonstrate your analytical skills to potential employers.

Let’s take a closer look at how these types of data stories are structured.

 

2. Structuring a Story

As with all stories, data stories comprise three parts: a beginning, a middle, and an end. Each part plays a different role in the overarching data narrative.

Figure 2. Every story needs a beginning, middle, and end—even data stories!

• Beginning: The beginning of the story needs to cover the “why” of your project. This should be the project motivation and objectives you outlined in your business requirements document. Think of it as any information that preceded and helped inform your project. It not only provides important context for your analysis but also sets the stage for what you’re about to present. Make sure to have a reminder for the project. Imagine being a CEO with a hundred different projects going on. Would you be able to remember all the aspects of each one?
• Middle: The middle of the story should cover the project plan and how you proceeded with your analysis. This should be the meat of your analysis: your hypothesis, your profiled data, and the statistical steps you took to analyze it. This part of the story will require some curation. Not every analysis you conducted needs to be included. Only incorporate those pieces relevant to the story’s beginning—and to your audience. You can do this by looking at each piece of information to determine whether it drives toward the eventual goal you want to make. If not, you might want to remove it from the story.
• End: The end of the story represents your conclusions and recommendations along with any contact information you need to add for future projects. What did you find? You may or may not have definitive conclusions or findings to present. You will, however, likely have some next steps or additional questions to ask. You want your narrative to lead your audience to these same conclusions or next steps. The conclusions you don’t find can be just as important as those you do! A lack of data can also determine what further research is needed.

The Structure of Harry Potter
To better imagine this story structure, let’s take a look at another story—the story of Harry Potter:

• The beginning is where you learn that there’s a wizarding world, that there’s both good and evil, and that Harry Potter plays a role in the fight against evil. This context sets the stage for the rest of the story.
• The middle of the story is where Harry and his friends learn how to be wizards and use what they’ve learned to fight against evil. They learn spells in school, make allies in the magical community, and face a number of trials. The story doesn’t cover every spell they learn and trial they face—this would create an incredibly long (and tedious!) story, indeed. Instead, it highlights the magical lessons and challenges that helped them grow and progress in their mission to stop the resurrection of the Death Eaters.
• The end is the story’s resolution. How did Harry and his friends use what they learned throughout the middle of the story to defeat Voldemort? Are there any questions that remain unanswered? What hints do you have from the story’s resolution on how the world will continue to grow and evolve beyond the end of the story as you know it?

 

3. Data Story Fundamentals

Before diving straight into creating your own storyboard in Tableau, let’s take a look at some fundamental approaches to data stories in general, as well as the impact they have as a method of communication.

 

3.1. Simplification

At their core, data stories simplify data, breaking it down into digestible pieces of information. They translate numbers, data, and analysis into relatable ideas that can be easily communicated to project stakeholders and casual viewers alike.

Take, for instance, the Canadian immigration example you looked at earlier:

Figure 3. The Canadian immigration example included again for your viewing pleasure.

This infographic poses a central research question: “Why is immigration important to Canada?” In fact, that very question literally points (in the form of an arrow) to the first data element—the fact that about 25 percent of the population will be over the age of 65 by 2035. This number is accompanied by a graphic of four people. One of those people is shaded to reinforce the sheer magnitude that is 25 percent of the population. This can be considered the beginning of the story, setting the scene for the data to come.

The visualization then uses a series of simple counts and rankings to guide viewers through a variety of different data sources—Canadian population data, world fertility data, immigration data, and more. These data points represent the middle of the story. Together, they help to answer the infographic’s central research question, with the accompanying graphics highlighting and simplifying the most important data elements.

The visualization ends by placing additional focus on the positive impact of immigrants. This provides a conclusion to the initial research question.

Similar to how you were taught to design your visualizations—as simple as possible—so, too, should your story seek to simplify and refine. Your project needs to be distilled down to its core components. While it should have a clear beginning (why), middle (analysis), and end (conclusion), they should only include the minimal information necessary to relay the story. This simplicity, together with the overarching narrative that ties everything together, makes analytical information more accessible to a broader audience. Analysis, at its core, is all about parsing large amounts of complex information into its simplest and most easy-to-understand version.

3.2. Engagement

In addition to simplifying information, data stories also improve audience engagement. Consider how you interact with social media (if at all). You’re likely attracted to content that’s a mixture of words and pictures. Marketing research supports this concept. Posts with visuals produce more clicks and, subsequently, more user engagement.

Using a combination of charts and text in your storyboards works the same way—it makes your work and research projects more engaging to stakeholders.

3.3. Retention

Have you ever heard people talk about being “left-brain dominant” or “right-brain dominant”? Left-brain dominance is usually connected with analytical thinking, while right-brain dominance is usually connected with creative thinking and is more visually oriented. Blending visuals with text targets both sides of the brain at the same time. The data and analytical components appeal to the analytical left brain, while the charts and pictures appeal to the visual right brain. As a plus, the narrative aspect engages your emotions. This “whole brain” experience increases retention of information. Data analysis involves just as much creative thinking as analytical thinking!

Figure 4. With their mixture of data, pictures, and narrative, data stories appeal to the entire brain!

Data stories map your journey from a research question to a final recommendation or conclusion, highlighting critical information along the way. Think of them as providing a framework to communicate insight and simplify the analytical aspects of your project.

Consider, for example, the weather forecast. Meteorologists work with complicated computer simulations to predict future temperatures and weather events. If they explained their detailed analysis as part of their weather forecasts, their audience would quickly shrink. Instead, meteorologists discuss the upcoming weather as a story—where’s the weather coming from, what locations will be affected, and how long will the weather last? Computer simulations supplement this narrative by showing the speed and path of the storm, adding a visual component to keep viewers engaged. This combination of elements makes for a more effective weather forecast than if the meteorologists simply provided tables showing wind speeds and pressures in surrounding cities.

 

Storytelling with Data

A key part of being a data analyst is communicating your analysis to other people, helping them grasp your findings and persuading them to react to your analysis in the right way. In other words, storytelling. In this piece we are going to cover where storytelling fits in the data analysis, process, how to read your audience, how to create a compelling narrative and how to back it up with the right visualizations. 

Storytelling often comes under the last step of data analytics process. You have collected the data, cleaned it, analyzed it, and have prepared visualizations. Now it's time to present these results to stakeholders.

Of course, often you need to present intermediate results. So whenever you need to communicate your findings to someone, you need the storytelling skills. And this brings me to the first out of the three key elements of storytelling: who is that someone that you need to present to. Who is your audience?

This can vary from members of the executive leadership team, to your technical peers from people in finance to those in sales or engineering or operations, etc., they may even be external partners or clients. Hence the first step of efficient storytelling is to know your audience, because this will allow you to number one, set the tone and number two, identify what message you want to communicate.

In short, it allows you to plan the layout of your presentation, your story, and you need to speak the language of your audience. If you get too technical with a non-tech audience, there is a risk that they won't understand you. On the other hand, if you get too non-tech with a technical audience, you risk losing the attention and credibility.

Each of these teams have different requirements and expectations. For example, your leadership team might want to know clear cut facts and patterns so that they can make decisions. On the other hand, your technical colleagues may be more interested in the raw data. Hence again, remember: know your audience.

Second key element is to construct a compelling narrative, which means setting the stage. Simon Sinek in his famous Ted talk "Start with Why" strongly advocates that one should always start with the why and purpose of their story. This helps in engaging your audience. We don't want to leave our audience hanging with the question: why am I told all this? What do I do with this data? We really need to set the expectations straight, right from the beginning. This can be done by posing a question that attaches to an overall goal or strategy.

The third key element is to organize your visualizations. Here, you need to be very careful. You have done a lot of analysis so far and have created a lot of visualizations, but all of that won't go into your story. You only pick and choose what makes sense. And what fits into the big picture. A common trap is to put lots of graphs on the slides, but we shouldn't do that.

A very good practice is to connect each visualization with a question. You ask a question and use the data that is your visualization to answer that question, and you can repeat this for each of your visualizations. So let's recap. Number one, you know your audience, number two, you set the stage by asking an inviting question that corresponds to a strategic decision or a bigger goal.

And then number three, you use your analysis and your visualizations to carefully lead your audience, step by step, as you address this question or goal. If you do all this right then by the end, they'll be on your side. And that's it. This is how you tell an effective story using data. 

A piece by Dr. Humera - Founder Lumen 

 

4. Storyboards in Tableau

Fortunately for you, Tableau comes with a built-in storyboard feature designed precisely for this storytelling approach. First, let’s walk through what one of these storyboards might look like for a data project. To follow along, open up this example project in your browser: Urbanization of China Storyboard

Along the top of this visualization are tiles describing the main story components—China’s Ambitious Goal, Urban Growth Rate Comparison, Urban Population of China’s Largest Cities, and more. In Tableau terms, these top tiles are the “story points.”

Figure 5. These top tiles are referred to as “story points” in Tableau.

This storyboard tells a narrative about China’s urbanization efforts. The first story point—China’s ambitious goal—is used to outline the beginning of the story, or the project premise. Namely, that China plans to move millions of rural residents to cities over the next 12 to 15 years:

Figure 6. The first story point represents the beginning of the data story and sets the stage for the rest of the narrative.

The story then engages the audience by relating the number “250 million” to city populations around the world. This provides a frame of reference, allowing audiences to understand the magnitude of this number.

There are clear directions on the screen for users to “use the slider to gain perspective on the number of equivalent major cities that represents.” Each click on the right-arrow of the slider increases the number of people and displays cities around the world with the corresponding populations. Always ensure you include instructions whenever viewers are supposed to interact with pieces of your storyboard!

Figure 7. By the time you hit 250 million, multiple cities are displayed. This is because no one city can match a population of 250 million!

This comparison of the data to well-known cities engages the audience and helps them understand the context of the project before diving into the actual analysis.

The next few story points walk through the data that describes this urbanization process. This would be considered the middle of the story. Each story point uses a combination of visuals and text to provide an overview of what the data and visualization show. Everything has been carefully curated to ensure that each piece aligns directly with the initial premise set out in the beginning of the story.

Figure 8. Not every analysis or graph needs to be shown. Only those that directly align with the initial premise should be included in the narrative.

The first supporting story point—Urban Growth Rate Comparison—shows the urban growth rate of China compared to other countries. To make this more relatable, this growth rate is compared to the growth rate of other countries. Likewise, how has this move towards cities changed the urban percentage over time for China? This, too, is compared to other countries, allowing viewers to consider the country’s current goal within the context of how its urban growth rate has fluctuated in the past.

Figure 9. This visualization makes it easy to see the population of each major city in China at a glance.

The next story point—Urban Population of China’s Largest Cities—focuses directly on China, examining the size of each of its existing cities. Imagine living in a city with a population of one billion. Would you notice if 250 million people moved into your city? What if the population of your city was one million? Would you notice if the population grew by a magnitude of 250 million then? This visualization includes every Chinese city that has a population greater than one million. The largest circle represents Shanghai, with a population of roughly 23 million. From this visualization, one can gather that the effects of 250 million new people would be felt by people in every city in China, even if those 250 million people were spread across multiple cities.

Figure 10. If China’s current urbanization trend continues, 250 million people will move into cities naturally by the year 2025.

This next story point—New Initiative or Current Trend?—examines how this policy interacts with China’s historic urbanization trends. China has already been urbanizing at an increasing rate. Is this policy even necessary? This visual presents an analysis predicting that the goal of moving 250 million people (or reaching a 70 percent urban population) would naturally occur in the year 2025 without government intervention (assuming past urbanization trends continued). This chart attempts to put the goal in greater context and set up the query posed in the next story point.

Figure 11. To expedite urbanization, the government is seizing rural land from farmers to make more room for cities. The government makes quite a profit from these transactions.

This next story point—Forced Land Takings by the State—queries the urbanization goal. If China is already on an urbanization trajectory, why is the government intervening? Well, according to the data, the government profits when relocating farmers and seizing their land.

In this way, the middle of the story puts the initial goal in context in terms of China’s current urban population and historic urbanization trends. It shows how many people need to move to achieve this goal, as well as how this relates to current movement trends. It then examines the motivation for this goal in terms of land acquisition by the government.

Figure 12. The leftmost graph shows the percentage of farmers who have lost some of their land, while the rightmost graph shows the percentage of farmers who have received access to urban schools. 

The end of the story comes in the form of the final story point, The Result. This point explores the two big tradeoffs (farmers’ land losses and farmers’ quality-of-life improvements through urbanization) in this goal of urbanization.

The story starts with China’s plan to relocate millions of rural residents to cities. It then examines the magnitude of the relocation efforts, urbanization trends in China and worldwide, as well as the impact on relocated rural residents. It finishes up by turning the focus to an individual level, highlighting both the positive (better access to schools and medical coverage) and negative (land loss) impact to the farmers.

Importantly, this ending doesn’t include a recommendation; rather, it provides unbiased comparisons between the land farmers lose and the gains they make through urbanization. This demonstrates that not all stories end in a clear recommendation or next step—some simply guide viewers through the data, leaving the conclusion open for interpretation. The beginning of the story opened with China’s urbanization goal, but it just as easily could have opened with a more pointed question: “Is urbanization good for rural citizens?” By taking this broader approach, it engages a larger base of users before narrowing its focus to farmers and rural residents. Neither beginning is better than the other; they simply serve different purposes.

 

5. Making a Tableau Storyboard

Throughout this Module, you’ve used a number of different data sources to demonstrate how and why to build visualizations in Tableau. In this section, you’ll bring everything you’ve learned together to create a storyboard using the ever-familiar candy data set. To do so, you’ll repurpose some of the visuals you’ve already made, as well as create new ones to ensure the story you’re telling is cohesive and has enough context.

Tableau allows you to create three different types of visualizations: sheets, dashboards, and storyboards. The icons for each are shown below, respectively, from left to right:

Figure 13. The three different types of visualizations you can create in Tableau are sheets, dashboards, and storyboards.

A storyboard in Tableau comprises a combination of sheets and dashboards:

• A sheet allows you to create a chart or visualization. You can choose the chart type or adjust the colors, sizes, text, and legends, the same as you’ve already been doing throughout this Module. Each sheet can only contain one chart.
• The middle icon designates a dashboard. You can’t create new charts on dashboards. Instead, you can arrange multiple charts—multiple sheets—onto a single page, creating a single visualization. You can add additional text and formatting when using a dashboard.
• The final option is a storyboard. This is where you can combine sheets and dashboards to develop a holistic data story. Compared to a dashboard, it has fewer layout options. Instead, the focus is on curating and arranging multiple sheets and dashboards.

 

Let’s walk through how these features were used to create the Chinese urbanization story above.

Figure 14. You can tell this is a dashboard because it contains multiple charts, visuals, and textual elements.

The feature used to create this page was the dashboard. You can tell this because it includes more than one chart or visual. The dashboard combines the urban population bar chart, which shows the change in urban population for each country between 1960 and 2013, together with a line chart that displays the corresponding trend line for each country. Clicking on a country in the left-hand chart (Brazil is selected in the example above) emphasizes that country’s line in the line chart and adds another text-based visual about that country underneath. This narrative component helps explain the charts and promotes the story-like feel of this visualization.

The ability to click on Brazil in the bar chart to highlight Brazil’s trend line in the line chart is an excellent example of Tableau’s interactivity.

This dashboard is made up of four individual sheets. There are three chart sheets—the urban population visual, the historic trends line chart, and the country map. A fourth sheet shows the text about the country. It’s not a chart; rather, a textual element that changes depending on which country is selected.

Downloading a Storyboard
You can actually download this dashboard from Tableau Public to further investigate each of its components. To do so, click the down arrow in the bottom-right corner of the storyboard, then select Tableau Workbook as the file format:

Figure 15. Simply click on the down arrow icon, then select Tableau Workbook to download the file in a format you can open in Tableau.

Upon opening the file, right-click the Urbanization of China storyboard tab and select Unhide All Sheets. You’ll then be able to interact with all the components and see how the storyboard was built—how the individual dashboards make up the storyboard and the individual sheets make up the dashboards.

Figure 16. Once you’ve opened the file in Tableau, right-click the Urbanization of China storyboard tab and select Unhide All Sheets from the dropdown menu.

Now that you know how sheets and dashboards come together to build a storyboard, it’s time to practice creating your own! As mentioned earlier, you’ll start by creating a practice storyboard using the candy data set you’ve used a few times already throughout this Module. Get started by opening up your example project from Lesson 6: Statistical Visualizations: Scatterplots & Bubble Charts. This should be the file with your bubble charts and scatterplots for the candy data set.

As a brief reminder, this data set includes one row for each type of candy. It was created to examine common Halloween candies and determine which one people prefer. Each candy has attributes designating whether it contains caramel, chocolate, or fruit, among others. Additional variables include Pricescale, which represents where the candy falls in price relative to the rest of the candy in the data set; Sugarpercent, which represents the percentage of sugar in each candy; and Winpercent, which represents the percentage that particularly candy won in the one-to-one candy matches. You’ve already looked at many of these variables in earlier Lessons. In this Lesson, you’ll bring everything together to tell a story about Halloween candy in a Tableau storyboard.

This story is going to need a beginning, a middle, and an end. It can be helpful to think through what you want this structure to be before building it in Tableau. One way to do this is by writing a story outline. Toggle through each visualization you’ve already created and determine how each relates to your story outline. Are there any visualizations you don’t need? Are there any visualizations that would be helpful to have that you haven’t created yet?

In the candy example:

• The beginning sets up the motivation for the analysis. Why are some candies preferable to others?
• The middle covers the analysis itself. Is preference based on price, sugar content, or candy manufacturer?
• The end brings everything together by way of a conclusion or recommendation. Where does the analysis point? Do the variables you examined influence preference? If so, what data do you have to support this? Or, if no relationship could be found between the variables and preference, what does this mean? What impacts candy preference? (Hint: it’s likely individual taste!)

 

5.1. The Beginning

As you’ve now learned, the beginning of a data story is an excellent place to frame your initial question and project premise. Remember to show the “why” and the motivation. In the China example, the first page of the story stated China’s urbanization goal and provided context as to how this magnitude compared with populations of other world cities. In the Halloween example, this is where you’ll look at candy rankings and frame the story goal of exploring why certain candies were preferred over other candies. You haven’t created any Tableau visualizations for this winning data yet, so you’ll need to start with the most basic creation unit in Tableau—the sheet.

If you’ve opened up your workbook from Lesson 6, you should already be connected to the candy data set, and all your variables should be correctly categorized. With that out of the way, you can jump straight to creating a new sheet. Name it something descriptive; for example, “The Winners.” The purpose of this sheet is simply going to be to display all the candies according to popularity; in other words, which candies won the most in the one-to-one candy matchups.

Drag the Candy variable from the Dimensions list to the Rows shelf, then drag the Winpercent variable from the Measures list to the Text box on your Marks card. This will create a table displaying the sums of the Winpercent variable:

Figure 17. You should now have an unorganized list displaying win percentages for each type of candy.

Format the table so that the winning percentages look like percentages (without any decimal points). Remember—the data itself is already stored as a percentage; however, you need to tell Tableau to display it as one by adding the “%” suffix and taking away the decimal points.

If you need a recap on how to do this, head back to Lesson 5: Statistical Visualizations: Histograms & Box Plots, where you created your histogram.

You want the most popular candies to show at the top of the table. This means you need to sort the candy according to win percentage. Click on the down arrow next to the Candy variable on the Rows shelf and choose Sort. On the Sort modal that pops up, change Sort By to Field, select Descending, then select Winpercent for the Field Name:

Figure 18. Sorting the table by the Winpercent variable will list all the entries according to how many times they won in the matchups.

Another thing you can do is add numbers to the rows, making it easy to see the exact popularity rank of each candy. Tableau doesn’t have an automatic row number feature; however, it’s still relatively easy to implement by way of a field.

Create a new calculated field to calculate the rank of the candy by win percentage (to do this, go to the Analysis menu and click Create Calculated Field). Name this calculated field “Candy Rank.” Then, invoke the RANK_DENSE() function. The RANK_DENSE() function takes an expression, as well as an optional argument telling Tableau to list things in ascending or descending order. The expression is what you’re already showing on the Marks Card—SUM(Winpercent). The optional argument in this case will be desc. This is because you want your list to be in descending order. Your final calculation should look like this:

RANK_DENSE(sum([Winpercent]),'desc')

Figure 19. Remember that you can always read more about a function on the right-hand side of the Create Calculated Field modal.

After creating your Candy Rank field, drag it to the Rows shelf, then change it to Discrete (as opposed to Continuous) using the dropdown menu. This keeps your visualization in the form of a table rather than a graph. Move the Candy Rank field to the left of the Candy variable (in your Rows shelf) so it’s the first column to appear in the table view:

Figure 20. Reordering the variables in your Rows shelf will change the order in which they appear within the table view.

As it’s pretty clear what each column represents, you don’t need to include any labels. Go ahead and turn them off by right-clicking any column title and selecting Hide Field Labels for Row. Notice also that the Candy and Percentage columns have a unique shading pattern—every other row is shaded a light gray color, which makes the table easier to read. This is great! However, this shading doesn’t carry over into your new Candy Rank column. You can fix this by right-clicking one of the cells in the table and selecting Format. In the menu that appears to the left of your Marks card, choose the Shading option by clicking the paint can icon at the top of the menu:

Figure 21. The paint can icon brings up a menu for shading the rows in your table. (Click Image to Zoom)

This will bring up a Row Banding dialog. Here, change the Level slider all the way to the left:

Figure 22. Use the Level slider to indicate how far across your table you want the shading to extend.

Immediately, your table should update, and the shading should extend across every column in your table:

Figure 23. Voilà! A fully shaded table!

With that, your introductory chart is finished. Now, you just need to start building it into a story. This first story point usually includes a chart with some kind of text; however, there’s no way to add extra text to a sheet. To do so, you’ll need to create your first dashboard, which can combine visualizations and text. This dashboard will serve as the introductory page of the story—in other words, the first story point.

Create a new dashboard in your workbook by selecting the middle icon from the bottom menu in Tableau:

Figure 24. The middle icon creates a new dashboard in your workbook.

Give it a descriptive title—for instance, “Beginning.” This title works well in this example because you’re practicing creating the different parts of a story. In an actual storyboard, however, the title should be more reflective of the content itself, for instance, “Which Halloween candy is the best?”

The left-hand panel allows you to adjust the dashboard size. Choosing the right size can be a bit complicated. If you’re designing a story for an audience at work, and you know they’ll be using desktops and laptops, you can choose a size that will accommodate those screens. However, if you don’t know who your exact audience will be, choosing a size can be difficult. That’s because people might be viewing your storyboard on any number of different devices, each with different sizes. You can either choose a common size or simply select Automatic, which will tell Tableau to make the storyboard look good on any screen. Regardless of what you choose, you should always test your final storyboard on multiple devices before sharing it publicly.

Figure 25. You can either choose a fixed size or let Tableau adjust the size of your storyboard automatically. 

First, you want your dashboard to display the table of candy winners. Locate your sheet, The Winners, listed under Sheets in the left-hand panel of your view and drag it onto the dashboard:

Figure 26. Adding a sheet to a dashboard is as simple as dragging it into your view! 

Right-click the title of The Winners on your dashboard and select Hide Title. You’ll be adding new descriptive text to replace this shortly. Keep this in mind when you’re creating storyboards. Since your individual sheets are no longer standalone visualizations, it’s okay if they don’t include as much descriptive text. You’ll be adding text on your dashboard to make up for it.

Since this will be the first story point of your storyboard, let’s add a graphic to make it more appealing. In the left-hand panel, Tableau displays a list of available objects you can add (under Objects). Go ahead and drag the image label onto the Tableau dashboard view. A box will appear (which will hold the image). You can drag this box around to move the image on your dashboard. For now, place it to the right of your table. Once finished, load the actual image you want to display within this box. To do so, right-click the image box and select Edit Image. This will trigger a modal asking you to provide an image:

Figure 27. The Edit Image modal allows you to select an image and choose from a small selection of display options.

Click the Choose button to bring up a file selector. For now, you can use this candy battle image. Save it to your device and upload it here as a PNG:

Figure 28. Remember this picture?

You can also add alt text to your image. This is text that will be read aloud if the viewer is using a screen reader, making this a great accessibility feature and one you should always try to include in your storyboards. Effective alt text should be a clear description of what the image contains.

Remember learning about accessibility back in Lesson 2: Visual Design Basics & Tableau? Reread the section on “Accessibility” for a recap on this topic.

Figure 29. Choose Fit Image to make your image the same size as the box you’ve already positioned on your dashboard, then add some descriptive alt text to your image.

Once finished with your image, drag a Text Object from the Objects area in the left-hand panel to the very top of your dashboard. This text will provide the project overview and goal. Something like the following would work well:

Halloween Candy Showdown

• 82 candies go head to head.
• 8,371 people vote.

The votes are tallied to find the candy with the highest percentage of winning battles.

With that, all of your pieces are in place! All that’s left is to apply some of the design principles you learned about at the beginning of this Module to ensure you’re using color and whitespace effectively. Try out some different formatting options until you find something you like. Feel free to use some creativity, as well! Halloween candy itself is a pretty playful topic, so you can have some fun with it!

One tip for making things look better is to adjust the fit of your visualizations. What does this mean? Well, each of the elements on your dashboard has an invisible box that it sits inside. The content of that box doesn’t always expand to fill up the entirety of that box, which makes for some awkward whitespace. You can tell a visualization to take up the entire width of its box by clicking the arrow on the right-side of the sheet. From the menu that pops up, select Fit→Fit Width:

Figure 30. The Fit Width option tells a visualization to extend horizontally, taking up the entire width of its containing box. You can do the same thing vertically by setting it to Fit Height. 

See how the rows of your table have extended all the way to the edge of the box? This looks considerably more visually appealing. Take some time now to adjust the size, colors, and whitespace on your dashboard, as well as add some more text. Again, feel free to be creative! Make your dashboard unique to you and your project! If you need some ideas, feel free to use the images below as guidance.

Figure 31. A Halloween theme of orange and black would be appropriately quirky!

The final addition to this first story point alludes to the coming analysis and overall research question. What makes a candy a winner? Which candy would you pick? That is where the middle of the story can take over.

With your beginning dashboard finished, you can add it to a storyboard. But first, you need to create that storyboard! You can do so via the third option in Tableau’s bottom menu:

Figure 32. The third option is the ever-elusive storyboard.

Give the story a descriptive, engaging title for instance—“A Halloween Tale.” The title is just meant to convey the broad subject area and catch a user’s eye. It doesn’t need to be as descriptive as the titles you created for your individual charts.

The dashboard you just created will become the first page in your story. Just as you dragged your “The Winners” sheet into the dashboard view to create your dashboard, you now need to drag your “Beginning” dashboard into the story view to create your storyboard:

Figure 33. You’ll always be dragging smaller elements onto larger elements, for instance, sheets onto dashboards and dashboards onto storyboards.

At the top of your screen, you can add a caption by clicking on the text and typing something—for instance, “Winning Halloween Candies.” This will become the tabs, or the story points, you saw earlier along the top of the China storyboard:

Figure 34. Think of these story points as the “menu” of your storyboard.

In the top menu, select Story→Format, which will bring up the formatting options for the story and allow you to change the font and shading to match the look and feel of your storyboard.

Figure 35. You can change the font and colors of your story points via the Format menu.

Right-click the title at the top of the storyboard and select Hide Title to hide the storyboard title at the top:

Figure 36. That’s one “sweet” storyboard so far!

This first dashboard explains the “why” of the story—in other words, its motivation or objective. It examines which Halloween candy people picked the most in head-to-head voting and provides a list of all the candies so users can see where their favorite candies rank. You’ve now set up the rest of your story, which will look into what exactly elevates a candy from “just okay” to “the epitome of candy excellence.” Onward!

 

5.2. The Middle: The Analysis

The middle is where your analysis lives. Remember, however, that you don’t need to include every analysis you conducted throughout the project when presenting your story. Instead, include only those analyses that directly support your story’s objective (i.e., what you presented in the first story point).

For a storyboard of your Cliqz project, for instance, you can reuse the visuals you created throughout Module and the data points you identified. Remember that you can’t directly copy sheets in Tableau Public, so you’ll need to recreate any graphs saved in a different file. Opening up two copies of Tableau will help you compare what you did previously side by side. And the extra practice should help solidify your learning!

The Halloween candy project introduced in this Module was less defined than your Cliqz project, so you’ll need to tweak some of your visuals to align with the story you want to tell and the question you want to answer (i.e., what makes a winning candy).

The three variables you’ll be exploring are Pricescale, Sugarpercent, and Manufacturer, along with the overarching question of whether any of these variables have an impact on candy popularity. In your Tableau workbook from Lesson 6 (the one you’ve already been working on in this Lesson), you looked at the correlation between sugar content and price. Copy this scatterplot twice and simply change one of the variables so that each chart looks at a different correlation: one between Winpercent and Sugarpercent and one between Winpercent and Pricescale:

Figure 37. Each of these charts will help paint a picture surrounding what variables affect candy preference. 

In Lesson 3: Comparison & Composition Charts, you created a treemap using this same candy data. You can incorporate that into your story’s middle, as well (with some tweaks). As Tableau Public doesn’t allow you to copy-paste, go ahead and recreate it.

The original treemap visualized candy according to manufacturer and hardness. In your new copy, replace the count of candies variable (CNT(Candy)), which is currently determining the Size and Text of the treemap, with the average of the Winpercent variable. While you’re at it, go ahead and color the treemap by the average of Winpercent, too, by adding it to the Color box on your Marks card:

Figure 38. Remember that you can replace variables by simply dragging a new variable on top of the old variable. 

From here, you may want to make a few extra aesthetic changes; for instance, playing with the color scale and changing the number format of Winpercent to be a percentage if you haven’t already done so. Your final treemap should look something like this:

Figure 39. Nothing like a good treemap to help “grow” your storyboard. 

Create a new dashboard and add all three of these visualizations. Each one explores a different metric that could be the reason people chose a certain candy as their favorite.

Figure 40. So many candy charts, so little time.

One thing to note here is that while your scatterplots examine individual candies, the treemap examines individual manufacturers. In other words, your visualizations are at a different grain. This is totally fine! However, you need to call attention to this in your dashboard so that viewers don’t get confused. This can be fixed with a few descriptive titles.

Additionally, since the treemap looks at manufacturers, which is a higher grain, you may want to add a count of how many candies each manufacturer makes (since this is the grain of the scatterplots). To do so, drag the Candy variable to the Label box on your Marks card. You’ll then have to select Measure→Count to display the count rather than the candy name. You may also need to reformat and tweak the text to ensure it’s easy to see that one number represents the win percentage while the other number represents the count of candies. To do so, click on the Text box on your Marks card and select the ellipses “…” to the right of the Text box to edit the text.

Figure 41. Edit the text by clicking on the ellipses next to the Text field.

Figure 42. Titles describing the grain of each visualization help viewers better understand and absorb the information.

One of Tableau’s strengths is its interactivity. This means that you can click on something and the visualization will change. In the growth rate comparison story point back in the China storyboard example, you could click on Brazil to change the visualization. Two things would happen: 1) text and a map of Brazil would appear at the bottom of the screen, and 2) the trend line for Brazil in the rightmost chart would be highlighted. All the sheets within the dashboard were connected:

Figure 43. Clicking on Brazil would change the visualizations on the dashboard to give you more-specific information.

Fortunately, Tableau makes this connectivity easy to achieve. Select a visual in your dashboard view to display a vertical menu on the right-hand side of the sheet. The down arrow at the bottom of the menu will bring up a menu with more options. From here, select Use as Filter. Do this for each of your visualizations within the same dashboard.

Figure 44. Setting Use as Filter for a visualization will allow you to set specific points as filters for other visualizations.

Now, if you click on a rectangle in your treemap, the two scatterplots will filter themselves based on what you’ve selected:

Figure 45. Clicking on a single rectangle will filter the other visualizations to only display data related to that rectangle.

Now, you need to think about the narrative component of your storyboard. Do you want users to be able to look for a specific candy? You probably want them to examine this candy in relation to the other types of candy. To do so, you can implement a highlight feature similar to what you saw in the China storyboard. Currently, your scatterplots only display Winpercent compared to Pricescale and Sugarpercent. If you want to highlight according to the type of candy, you’ll need to add this variable to the scatterplots, as well. You can do this by dragging the Candy variable to the Details box on your Marks card for both individual scatterplot sheets.

Return to your dashboard and select one of the scatterplots, then bring up the options menu by clicking the down arrow on the right-hand side. Select Highlighters→Candy:

Figure 46. Highlighting lets a viewer “highlight,” or focus on, a specific data point.

This should create a Highlight menu in the top-right corner of the dashboard:

Figure 47. From this new highlight menu, you can choose a specific type of candy, which will automatically filter all three visualizations.

This highlight feature allows users to view their favorite candy on the dashboard. It’s a bit hard to see in the bottom-right hand corner, so go ahead and move it to a more prominent spot, such as the top of the dashboard. You can also add some prompt text asking users to pick their favorite candy:

Figure 48. The prompt will encourage users to interact with your storyboard and learn more about their favorite candies.

From here, think even more about the story you’re trying to tell. How can you best arrange the pieces to explain your analysis and keep viewers engaged with the data? This often involves adding text and making titles more descriptive. Pretend you’re talking to a user. What would you tell them about the visual they’re seeing? What would you want them to click on and play around with?

Annotate points to draw attention to certain data items. The candy data doesn’t have any high correlations (sugar content, price, nor manufacturer had much of an impact on win percentage). Use an annotated point to drive that home. Find the winning candy and check whether it’s the cheapest. Point this out to users with text. To do this, right-click the data point in the corresponding sheet and select Annotate→Mark, then add any key information you want to include alongside the data point:

Figure 49. Add further information to key data points with annotations!

Finally, format your dashboard and add it as a new story point to your storyboard by selecting Blank under the New Story Point up in the top-left menu under the Story tab:

Figure 50. If you don’t see the Story tab, check that you have your storyboard open (and not your dashboard or one of your sheets).

Give your story point a descriptive caption!

Figure 51. Your storyboard should now have two story points!

 

5.3. The End: Recommendations, Conclusions, and Next Steps

This quick analysis didn’t find any concrete evidence that price, sugar content, or manufacturer has any impact on the popularity of a candy. However, your story still needs an end. You can choose to show this lack of conclusion in a variety of manners. If your middle is long and complicated, for instance, you may want to explicitly remind users of your project’s motivations. Because this example only had one middle story point, however, this shouldn’t be necessary—your viewers will likely still remember the objective from your first story point.

Figure 52. Your analysis made it clear that price, sugar content, and manufacturer don’t have any impact on candy preference. From this, you can surmise that candy preference is simply a matter of taste!

The end of your story is where you summarize the analysis for your audience. You explored a hypothesis and examined data in different ways. The ending, then, should tell your audience what the next steps are and/or what the analysis discovered.

This example project chose to examine candy preference according to price, sugar content, and manufacturer. However, none of these variables explained why certain candies were preferred over others. This last dashboard should simply reiterate those same ideas. As the story itself is simple, its ending can be simple, too.

This last dashboard should summarize your analysis by showing the most and least popular candies along with their price, sugar content, and manufacturer information. These are the same rankings shown in the first story point and the same three variables analyzed in the middle story point—you’re simply displaying them in a different way so you can make the case for your conclusion. The image in Figure 53 below is just one approach you can take for this final story point, but you can design it however you like!

Figure 53. Price, sugar content, and manufacturer are summarized through a series of charts that make it clear no one variable has a large impact on preference.

This story may go on to have more chapters that look at additional data points—or it may end here. Just remember that a finding of no significance is still a valid, conclusive finding. That being said, you still want to leave your audience with an action from your findings. That’s why you’re doing an analysis in the first place! In this case, that might be directing what future research is needed, for instance, looking for data other than price, sugar content, or manufacturer. What other ingredients are in the candy? Peanuts? Nougat? Chocolate? Does data exist on those statistics?

If your results do find a relationship between variables, then you’ll have a more structured conclusion, addressing each of the motivational points that led to your investigation. Especially if you won’t be present to show off your analysis, be sure the whole story can stand on its own without any need for further explanation. You might need to include contact information so your audience can find you to ask questions.