some visual features we can use to display data effectively. Now we need to look at the different types of data, in order to choose the best visual encoding for each type.
Types of Data
There are three types of data: categorical, ordinal, and quantitative. Let's use a photo to help us define each type.
Categorical Data
Categorical (or nominal) data represents things. These things are mutually exclusive labels without any numerical value. What nominal data can we use to describe the gentleman with me in the Figure 1.11?
● His name is Brent Spiner.
● By profession he is an actor.
● He played the character Data in the TV show Star Trek: The Next Generation.
Figure 1.11 One of your authors (Andy, on the right) with a celebrity.
Source: Author's photograph
Name, profession, character, and TV show are all categorical data types. Other examples include gender, product category, city, and customer segment.
Ordinal Data
Ordinal data is similar to categorical data, except it has a clear order. Referring to Brent Spiner:
● Brent Spiner's date of birth is Wednesday, February 2, 1949.
● He appeared in all seven seasons of Star Trek: The Next Generation.
● Data's rank was lieutenant commander.
● Data was the fifth of six androids made by Dr. Noonien Soong.
Other types of ordinal data include education experience, satisfaction level, and salary bands in an organization. Although ordinal values often have numbers associated with them, the interval between those values is arbitrary. For example, the difference in an organization between pay scales 1 and 2 might be very different from that between pay scales 4 and 6.
Quantitative Data
Quantitative data is the numbers. Quantitative (or numerical) data is data that can be measured and aggregated.
● Brent Spiner's date of birth is Wednesday, February 2, 1949.
● His height is 5 ft 9 in (180 cm) tall.
● He made 177 appearances in episodes of Star Trek.
● Data's positronic brain is capable of 60 trillion operations per second.
You'll have noticed that date of birth appears in both ordinal and quantitative data types. Time is unusual in that it can be both. In Chapter 31, we look in detail about how you treat time influences your choice of visualization types.
Other types of quantitative measures include sales, profit, exam scores, pageviews, and number of patients in a hospital.
Quantitative data can be expressed in two ways: as discrete or continuous data. Discrete data is presented at predefined, exact points – there's no “in between.” For example, Brent Spiner appeared in 177 episodes of Star Trek; he couldn't have appeared in 177.5 episodes. Continuous data allows for the “in between,” as there is an infinite number of possible intermediate values. For example, Brent Spiner grew to a height of 5 ft 9 in but at one point in his life he was 4 ft 7.5 in tall.
Encoding Data in Charts
We've now looked at preattentive attributes and the three types of data. It's time to see how to combine that knowledge into building charts. Let's look at some charts and see how they encode the different types of data. Sticking with Star Trek, Figure 1.12 shows the IMDB.com ratings of every episode of Star Trek: The Next Generation.
Figure 1.12 Every episode of Star Trek: The Next Generation rated.
Source: IMDB.com
Table 1.3 shows the different types of data, what type it is, and how it's been encoded.
Table 1.3 Data used in Figure 1.12.
Let's look at a few more charts to see how preattentive features have been used. Figure 1.13 is from The Economist. Look at each chart and see if you can work out which types of data are being graphed and how they are being encoded.
Figure 1.13 “A terrible record” from The Economist, July 2016.
Source: START, University of Maryland. The Economist, http://tabsoft.co/2agK3if
Table 1.4 shows how each data type is encoded.
Table 1.4 Data used in the bar chart in Figure 1.13.
Let's look at another example. Figure 1.14 was part of the Makeover Monday project run by Andy Cotgreave and Andy Kriebel throughout 2016. This entry was by Dan Harrison. It takes data on malaria deaths from the World Health Organization. Table 1.5 describes the data used in the chart.
Table 1.5 Data used in the bar chart in Figure 1.14.
Figure 1.14 Deaths from malaria, 2000–2014.
Source: World Health Organization. Chart part of the Makeover Monday project
How did you do? As you progress through the book, stop and analyze some of the views in the scenarios: Think about which data types are being used and how they have been encoded.
Color
Color is one of the most important things to understand in data visualization and frequently is misused. You should not use color just to spice up a boring visualization. In fact, many great data visualizations don't use color at all and are informative and beautiful.
