A journalist or blogger who uses Tableau Public must convert data to visualizations so that readers can quickly understand large quantities of data that have been distilled down to a single graphic. The role of an author is data abstraction, the conversion of real-life data into visual cues (primitives, as some people who work in the field of data science call them), such as colors, shapes, lengths, and positions. The role of your reader is to do the reverse, consume data primitives and convert them back into real-life things, events, and phenomenon. If you have done a poor job choosing the elements of your data visualization, your readers will have a difficult time relating your visualization to the real world, which means that the story that you're telling isn't as compelling as it should be. Your visualization will not have the value that you wanted, because the connection between the data visualization and the underlying data has been broken.
Before you create a visualization for people to utilize, there are several considerations that go into the design. The following are some of these considerations:
- The purpose of the visualization: Which questions should it answer?
- Content: What should be included? (Start small, as less is often more).
- The visualization structure: What do you actually want to measure or show?
- The readers' platform: How have your readers approached your questions before? Which tools have they used to do so?
In this chapter, we will discuss the following topics:
- The lifecycle of visualization
- Iterative design and what it means
- Creating visualizations and visual perceptions
- Types of visualization
- Using discrete and continuous dates and measures
- Filtering, grouping, and sorting
Perform the following steps to develop a high-quality, compelling visualization:
- The first step is to devise a purpose for the visualization—a question that you would like the data to answer—and envision the possible visualization types that can best answer that question.
- Next, define the content and find your relevant data; determine how much data should be shown and the level of details required. A large part of your visualization project should be spent on finding and cleansing data. Do not skimp on this step, because it will make your visualization more clear and useful to readers. You may have to scrape data from the Internet or combine it from multiple sources.
- Wherever you get your data, make sure that you have both validated and attributed it properly.
- The third step (with respect to this book) is to try various view chart types using the Show Me feature of Tableau Public. Try various view types to answer the question, experiment with various views, add or subtract data, manipulate components, and explore the question in greater detail or find other questions.
- The next step in building a visualization is considering the platform—print, desktop, laptop, tablet, and smartphone. The medium that your consumers use to interact with your visualizations and develop insights is critical.
- The last step is publishing your work, or delivering content to your visualization consumers. Your data visualization should start a conversation. So expect feedback.