Top 5 Essentials for Your Data Visualization Portfolio

posted in: How-To's and Guides | 0
Top 5 Essentials for Your Data Visualization Portfolio | Seattle Central College - Continuing Education
Top 5 Essentials for Your Data Visualization Portfolio | Seattle Central College - Continuing Education
Top 5 Essentials for Your Data Visualization Portfolio | Seattle Central College - Continuing Education

Develop a Portfolio with Intention

Guest post by Jenny Richards

Students in my data visualization classes often ask me about the work in my portfolio — where I came up with ideas, how I got people to give me data, what’s my best work, etc.

While not everything in my portfolio was intentional, I do have a few visualizations that I think demonstrate specific skills, and it’s often those I share with potential customers and collaborators.

Sometimes the viz shows off data cleaning and shaping; other times it’s storytelling.

But I look at every project — big or small — as an opportunity to add another item to my portfolio, something that could come in handy in the most unexpected ways.

That said, building my portfolio has been a bit haphazard and certainly opportunistic.

Had someone suggested earlier what sort of skills to demonstrate in my data visualizations, I’d have looked for opportunities to learn and show off those skills intentionally.

Now that I’ve gone from student to coach, I’m taking this opportunity to suggest five data visualization skills you should demonstrate in your portfolio.

The topic doesn’t matter as much as your ability to show that you understand how and why these skills are applied.

With each of these, I offer a couple of example visualizations from the Tableau Public community that I think demonstrate these skills well, but you could use any software: The skill’s the thing.

Each skill I show relies on out-of-the-box functionality in Tableau, and in addition to learning these in my class, Data Visualization with Tableau, there are numerous videos, blog posts, and community forums that explain these, too.

Maps

Why maps? Generally, people love maps. They identify with places on the map, they can manipulate what they see by zooming in and out, and they often provide complementary information to other data on the same dashboard.

Maps are so easy in Tableau that people don’t bother to make them special. And that’s too bad because it’s not very hard to make a map really pop.

Here’s a map in Tableau without any customization, and the same map after a few simple changes. Both are using the same data, bicycle accidents in Seattle from 2000 – 2016:

Before

Top 5 Essentials for Your Data Visualization Portfolio | Seattle Central College - Continuing Education

After

Top 5 Essentials for Your Data Visualization Portfolio | Seattle Central College - Continuing Education

I made the following changes:

  • Added a Mapbox background map
  • Resized the marks so they’re smaller
  • Color-coded the types of accidents

That’s it. Not much, but what a difference that makes!

Example of Best Practice

Here’s a data visualization from Ann Jackson that expertly uses an interactive map as the background for a story about the tree census in New York City.

Top 5 Essentials for Your Data Visualization Portfolio | Seattle Central College - Continuing Education
Click on the image to engage with the fully interactive viz.

The map adds to the story about the trees — you can hover over the marks on the map and it will tell you the species, ID, and status.

Other terrific map examples:

Change Over Time

What is change over time? It’s exactly as it sounds: Looking at one thing over time to see how it’s changed.

You want some examples of things you can measure over time? How about the number of penalties a player commits each year, comparing year to year? Or the number of malaria cases in a village in Zambia over 30 years? Or the value of a certain stock symbol?

Measuring change over time is probably one of the first charts you built in Excel, and it should be one of the first skills you demonstrate in Tableau. A standard line chart is a change over time. But just like the map example, you can take a simple line chart and turn it into something else with a few changes.

Here’s what a simple line chart looks like in Tableau without any polish:

Top 5 Essentials for Your Data Visualization Portfolio | Seattle Central College - Continuing Education

Not very compelling, right?

Take time to make changes like these:

  • Show more than one thing so you can compare several items as they change over time
  • Add color, smartly
  • Change the axis labels, background colors, and grid lines
  • Change the line format
  • Consider a different chart type — maybe a bar chart might read differently (see the History of the NFL below, as an example)
Example of Best Practice

Curtis Harris, a Tableau Public Ambassador, and Iron Viz Champion, has one of the finest examples of change over time in his History of the Single Season Home Run Record viz.

Top 5 Essentials for Your Data Visualization Portfolio | Seattle Central College - Continuing Education
Click on the image to engage with the fully interactive viz.

Looking for more examples of change over time? Here you go:

Custom Shapes

Instead of dots, squares, and bars, why not bring something that adds more detail without ‘costing’ too much? I’m going to start with an example.

RJ Andrews created this visualization to show the status of animals in Africa.

Using custom shapes, he lets the reader explore the animal’s territory (the map updates based on the animal selected), its endangered status (color), and whether the animal population is increasing or decreasing (depending on the direction the animal is facing).

Top 5 Essentials for Your Data Visualization Portfolio | Seattle Central College - Continuing Education
Click on the image to engage with the fully interactive viz.

Yes, he could have built a map that had dots for each animal and colored the dots the same as he’s done above. But by using custom shapes, the animals became the focus of this work — not the map. Plus, by changing the animal’s orientation, he’s cleverly added another data point without adding anything else to the page.

Learn to use custom shapes, but also learn when to use them. It may seem novel to include logos, but does it add to what the data is saying, or are you doing it merely to dress up an otherwise straightforward (boring?) viz?

Country flags can also be made into custom shapes, but don’t forget to think about your audience — will they know the difference between those flags that have the same colors, but in a different order? Does using a custom shape add anything to your work?

There are some risks to using custom shapes, such as:

  • If the files are too large, it can make your visualization slow
  • If there are a lot of complex shapes, it can make your visualization hard to read
  • If you have limited space on the page in which to show a lot of data, custom shapes might not be the best thing for your viz.

And don’t forget — logos are often copyrighted material, as are a lot of non-logo images. If you’re going to use them, use them judiciously, and legally.

Here are some other examples of vizzes with custom shapes. There are some things in here I’d do differently, but they’re very creative:

Good Color Palette

Color is all about restraint. Well, that and the fact that when you’re looking at a lot of different marks with different colors, the brain has a tough time processing that much data with precision. Use color because it means something, or because you’ve selected an intentional color palette that you’ll use throughout the viz. Don’t use color as a crutch – to make an otherwise boring analysis ‘fancy.’ Color doesn’t make a boring analysis fancy, it just makes a boring analysis colorful, and sometimes busy.

Use color because it means something, or because you’ve selected an intentional color palette that you’ll use throughout the viz. Don’t use color as a crutch — to make an otherwise boring analysis ‘fancy.’ Color doesn’t make a boring analysis fancy, it just makes a boring analysis colorful, and sometimes busy.

Don’t use color as a crutch — to make an otherwise boring analysis ‘fancy.’ Color doesn’t make a boring analysis fancy, it just makes a boring analysis colorful, and sometimes busy.

Here’s an example of too much color without any reason for it. Can you tell the difference between the first dot in 2004 and the first in 2005? I can’t.

Top 5 Essentials for Your Data Visualization Portfolio | Seattle Central College - Continuing Education

Color is also cultural, so consider your selections carefully.

For example, look at the chart below. What do you think is being measured here? Perhaps something about the holidays? That’s what I’d assume because of the color selection.

Top 5 Essentials for Your Data Visualization Portfolio | Seattle Central College - Continuing Education

Example of Best Practice

Jonni Walker is as master of the color palette. In the below example, he takes 3 colors and tells an incredible story. There’s not much data here, but he uses layout, color, and restraint to his advantage.

Top 5 Essentials for Your Data Visualization Portfolio | Seattle Central College - Continuing Education
Click on the image to engage with the fully interactive viz.

I know you want to see more examples of great color palettes. Here you go:

Small Multiples

When you’ve made a good chart, why not use it over and over again?

This is the idea behind small multiples — using many of the same chart type to compare. Compare what? In this example from Shine Pulikathara, he’s comparing crime statistics:

Top 5 Essentials for Your Data Visualization Portfolio | Seattle Central College - Continuing Education
Click on the image to engage with the fully interactive viz.

There’s a lot to see all at once, and sometimes small multiples can be overwhelming. This is where the other rules apply: Exercise restraint with colors and shapes. You should consider small multiples a variation of change over time, but a sophisticated and complex approach to a lot of data.

Most of the small multiples I’ve seen are about change over decades to geographies, as shown in several examples below. That’s because longitudinal study data is often consistent and structured in a way that facilitates this sort of comparison.


If you can master the five skills I’ve shown, your portfolio will reflect what you’ve learned. Don’t stop with just these five, though — see if you can spot trends in the examples I’ve provided and add more skills to your portfolio!

Want to learn more? Join me in Data Visualization with Tableau and start learning how to transform data into meaningful visual stories.

Share

Leave a Reply