Last modified on February 5, 2020 by Brad Phillips

Public Speaking Report 8: 5 Ways To Design Better Data Slides

Eye with a futuristic look

You’re in a data-heavy industry or speaking to an audience that values data. But that doesn’t resign you to using slide after slide of dense, hard-to-read charts.

When you pick the right visual, clear the clutter, drop the nonessential information, and use proven design principles, you can reduce the cognitive load on your audience and help them more quickly interpret your data.

Your scientific findings, your company’s financials, or in-depth statistical analyses become more dynamic and memorable when your audience can better see the story your data tells.

In this report, we offer you five tips on how to better do that.

5 Ways to Better Display Complex Data

You know the message you want your audience to remember, and the data that will help you to share it. Now, you need the best way to show that.

1. Determine What You Want to Show

Data is just data until you give it context and meaning. Do you want to show …

A relationship between two or multiple data sets? (When gas prices go up, revenue drops.)
Trend over time? (The average global temperature has risen over the past 20 years.)
Comparison? (The return on investment was higher this year than last.)
Distribution? A pattern? (The volume of shares that are traded at a specific price.)
Individual parts that make up the whole? (Total sales broken down by each product you sell.)

Once you know the “story” of your data, you can pick the best visual aid to say it.

2. Select the Proper Visual

The visuals you choose can help your audience more quickly interpret the data.

One way you can do that is to employ what Andrew Abela, author of Advanced Presentations by Design, calls the “squint test.” If your audience were to squint, so the text became blurry, could they understand the point of your data? The design or layout of some graphs better convey comparison, for instance, while others better highlight how parts fit into the whole. The right layout will give a better sense of your data “story.”

As Cole Nussbaumer Knaflic notes in her book, Storytelling With Data: A Data Visualization Guide for Business Professionals:

“If you’re wondering ‘What is the right graph for my situation,’ the answer is always the same: whatever will be easiest for your audience to read.”

In this report, we focus on the more commonly used graphs for technical, scientific, and academic data, as well as tips on how best to use them.

Table

In general, tables are not the most effective of visual aids, as they are more difficult to visually interpret. They don’t provide the visual cues that other visuals do, such as differing bar lengths, a rising (or falling) trend line, or a curve that shows distribution. Still, you may need to or be required to use one.

When to Use Tables:

To present exact numerical data of two or more variables
To show comparisons, trends, relationships

Best Practices:

Drop heavy borders. Either lighten them or create minimal ones. (See table below.)

Eliminate data. Do not include material that doesn’t support the key point of your slide.

Round off numbers. All those decimals could become distracting.

Size it right. It should be large enough to see in the cheap seats.

Provide emphasis. Use bold font on column or row headers, or particular numbers, to provide relevance and establish data hierarchy.

Here’s an example:

Heatmap Table

This version of a table color codes certain cells of important data points. (There are more sophisticated heatmaps where cells are of varying color or saturation to show relationships, comparisons, or patterns within your data or changes over time.)

Best Practices (in addition to those for tables):

In addition to those listed for tables:

Emphasize with color. Choose one color and adjust the value to highlight cells significant to the point you are making. Perhaps you pick blue. A darker blue might indicate the highest value, while a lighter blue indicates the lowest. Also, make sure text contrasts well with background color and is not challenging for those who are color blind. (See table below.)

Here’s an example:

Line Graph

One of the more common graphs, the line graph analyzes continuous data (one or multiple categories) over time. For instance, you could track a range of temperatures over several days or show the fluctuation of a product’s price (or two products’ prices) over 10 years.

When to Use a Line Graph:

To track trends over time with single, double, or multiple series of continuous data (Bar graphs do this too, but it is easier to see the trend in a line graph.)
To plot out individual points
To compare data
To show distribution patterns

Best Practices:

Time it right. Time intervals should run from left to right. The baseline (horizontal line) must reflect equal and consistent intervals. For instance, don’t mix months and years.

Clearly label axes. Avoid vertical or diagonal labels, as they can be harder to read. Consider labeling your line (lines) directly on the chart, rather than providing a legend.

Remove the border and grid lines. If you keep grid lines, lighten them to fade into the background.

Limit the lines. Keep to no more than four lines for each chart. Otherwise, you have a spaghetti graph that often is hard to read and interpret. Use a different color for each line. (As with other times that you use color, ensure that the combination you choose will work for people with partial or total color blindness.)

Baseline basics. The baseline doesn’t have to begin at zero (which is a best practice for bar graphs). Perhaps you are starting at a certain date, such as the day of the first known case of a pandemic illness or the first day you started recording a stock price.

Similar Graphs

Area charts. These also are used to show how data changes over time. However, instead of a single line, the space between the line and baseline is filled with color. They may better show how parts of a whole compare – such as the annual revenue from two different departments.
Timelines

Here’s an example of a line graph:

Source: Pew Research Center

Bar graph

A bar graph uses horizontal or vertical bars (also known as a column graph) to compare data among categories over time or some other measure. For instance, you might want to reveal survey results among certain age groups, or you want to show a stock’s open, high, low, and close prices over a stretch of time.

When to Use a Bar Graph:

To track trends over time with single, double, or multiple series of continuous data
To compare data
To show distribution (parts of a whole)

There are two types: horizontal and vertical.

Horizontal

When Best to Use Them (vs. Vertical):

When you have lots of bars to plot. (Still, refrain from too many to avoid cognitive overload. Consider breaking up the results into multiple slides.)
When your category labels are long.

Source: Pew Research Center

Vertical

When Best to Use Them (vs. Horizontal):

When you only have a few categories.
When you want to show negative and positive values. (You can do it horizontally, too, but we tend to read it better vertically.)
When you are comparing a change of some data set over time. (It’s easier for us to asses that if presented from left to right rather than from top to bottom, or bottom to top.)

Here’s an example:

Student achievement at 123 schools participating in the National Science Foundation’s Math and Science Partnership program. Source: Nicolle Rager, National Science Foundation