Ann Jackson - Tableau Strategies

• YYYY-MM-DD: First Release

With Early Release ebooks, you get books in their earliest formthe authors raw and unedited content as they writeso you can take advantage of these technologies long before the official release of these titles.

This will be the 1st chapter of the final book. Please note that the GitHub repo will be made active later on.

If you have comments about how we might improve the content and/or examples in this book, or if you notice missing material within this chapter, please reach out to the authors at ann@jacksontwo.com and luke.stanke@tessellationconsulting.com.

Categorical analysis is the foundation of data visualization. It is the first and most frequent type of data visualization data analysts use. Categorical analysis takes a dimension (example: Regions) and breaks apart by a measure (example: Sales). A dimension is typically a categorical value; these do not get aggregated. They are likely used to create data headers or to generate filters. A measure is a (usually numerical) value that can be aggregated using mathematical functions (like sum, average, or median). Measures create unbroken axes, those that extend from one end of a range to the other.

This type of analysis aids in answering common business questions such as:

• How does A compare to B?
• How is X measure distributed across Y categories?
• How much do A, B, and C contribute to the total?
• How does X measure change over time (where time is the dimension)?

Categorical analysis is usually presented as bar charts. Bar charts use height or length as visual encoding to express a measure. Visual encoding refers to different techniques for displaying data in charts. Encoding data in bar charts is very effective because humans can quickly analyze the variation among the size of the bars; they are also very easy to understand and label.

Figure 1-1. This illustration shows the various ways that data can be encoded for display. It aligns them to a comprehension scale of how precise the human eye can discern differences.

In our first use case, we will explore how to make effective bar charts. Youll play the role of a large financial institution that wants to understand which merchant categories make up the majority of transactional spend in order to drive marketing efforts and partnerships and better serve customers interests. We will also expand from the defaults and learn two additional methods for making bar charts that demonstrate the most important information.

In the second case study, on treemaps, youll learn about working with many dimensions. While bar charts are very useful, there are other data visualization tools youll need to leverage when doing categorical analysis. Sometimes when there are many members of a dimension, it becomes problematic to display each member as a bar chart. When this happens, you can use alternative chart forms to conserve space but still display all members. The most useful chart for this scenario is a treemap. Here youll play a non-profit organization that controls and awards grant money for Creative, Performing, and Cultural Arts Programs and Initiatives for the state of New York.

In our final use case within the chapter, youll learn how to use pie charts and donut charts to visualize whole relationships. Here, you are conducting a survey about IT professionals, their employment, their mental health and their employers attitudes toward mental health. Often these are the first type of data visualization you learn in school, but we like to use them sparingly and as an alternative option. By the time we get to this use case, youll see how properly executed pie charts can be great tools to craft and share data with your audience.