# 6. Gradients, Colors, and Shapes

Overview

When creating any visualization, you need to understand how best to incorporate different traits of the visual hierarchy. This lesson will help you with this while completing the second case study.

To explore more Kubicle data literacy subjects, please refer to our full library.

Summary

1. Lesson Goal (00:12)

The goal of this lesson is to learn how to incorporate different traits from the hierarchy into this visualization.

2. Using the New Visualization (00:23)

Our final visualization actually contains three separate visualizations: the gradient matrix showing sales by product categories and quarter, then two bar charts showing total sales by category and total sales by quarter. These three visuals work together as a single visual concept.

In our case, the three questions we wanted to answer from our stacked bar chart can be answered from the three new visualizations we created. Each visualization answers one of the questions, and does so in an effective and intuitive way.

3. Challenges with Using Gradients (02:03)

One issue that can arise with gradients occurs if a small number of values are larger than the bulk of the data. This can distort the scale used in the gradient. One way to deal with this os to cap the maximum range used in the gradient, so that all values above the cap have the same shading.

Another alternative is to use categorical values instead of a continuous scale. In this case, you would replace the gradient with distinct colors. This allows similar values to be identified easily.

4. Replacing Color with Shape (02:20)

If using categorical data, you may want to use shapes instead of colors for each category. However, this usually makes visualizations more difficult to understand, even if the number of categories is relatively small.

Transcript

In the previous lesson, we overhauled a stacked bar chart with the aim of creating a better visualization.

In this lesson, we'll learn how to incorporate different traits from the hierarchy into this visualization.

Let's now finish reinventing the stacked bar chart. When we broke apart the original chart, the software sorted the categories on the left alphabetically. But now, we can sort, for example, by total sales for each category. Better yet, you can add that information which was not in the original stacked bar chart. We can also add back the quarterly totals, and it's easier to follow than from stacked bars.

This is three visuals working together as one visual concept.

To recap, we started with the questions to answer and then used the hierarchy to explore options. Let's remind ourselves of those questions and see how this visualization answers them. First, we wanted to know, what is the trend? We can answer this by looking at the bars at the top of the chart, which tell us that sales seem to be increasing in 2019.

Second, we wanted to know which category had the highest sales.

This is answered by the bars on the left, where we can see that storage has the highest sales.

Finally, we wanted to know which quarter had the highest and lowest sales for each category.

We can easily identify these, as they're the quarters where the numbers are. So as we already saw, four categories had their lowest sales in Q1, 2016, and four categories had their highest sales in Q4, 2019.

As we can see, all our questions can be easily answered with this improved visualization.

This is a powerful technique for creating visuals that are not only easier to read but are more meaningful.

Now, let's take this gradient view through the rest of the hierarchy and discuss one key challenge with gradients.

You can easily locate the top few values, but you don't really get a good sense of the range of values because there are a few green squares and the rest are gray or white. If you have a set of numbers where only a few are at the bottom or top end, that skews the gradient. Here, the range is capped so that anything over 20,000 is dark green and you can see more of the higher values.

Let's say, instead of knowing the range of values, it's enough to identify high, medium, and low values. That's a perfect use for color.

Think of high, medium, and low as categories, not a range of values. It's now easy to see patterns forming in terms of product sales, and when they're selling best.

To finish this example, let's go all the way through the hierarchy and try shapes instead of color.

One thing to keep in mind when using shapes is to associate meaning as much as possible, like before, with up and down arrows for positive and negative numbers. That way, your brain doesn't have to work as hard to interpret the shapes and remember relationships.

Let's pause and think about what just happened. We flipped from color to shapes, and the view is much harder to process, and that's with only three shapes. This is where understanding the hierarchy really helps you come up with options.

We've now reached the end of the second case study in this course. As we've seen, understanding the visual hierarchy may lead to you creating charts that look quite different from software defaults, but they can actually convey information a lot more effectively than the defaults.

In the next lesson, we'll move onto our third and final case study where we'll overhaul a table of data.

Understanding Data Visualization
Applying Visual Data Skills

Contents

03:21

05:12

04:02

03:56

04:02

04:17

04:01

02:29

03:11

#### 10. Identifying Issues on a Dashboard

04:10

My Notes

You can take notes as you view lessons.

Sign in or start a free trial to avail of this feature.