Tricks for SAS Visual Analytics Report Builders: Bonus - Chart Best Practices

This is the third part of a six part series that describes tips and tricks for building impactful reports in SAS Visual Analytics.

Draft a Plan
Choose the Best Chart
Focus on What’s Important
Consider the Layout
Test, Test, and Test Again

SAS Visual Analytics enables you to create compelling, interactive reports that can be viewed by anyone, anywhere. To create impactful reports that resonate with your audience you need to (1) draft a plan for the report, (2) choose the best chart type to display your data, (3) create your reports so viewers can focus on what’s important to them, (4) pick a layout that will best display your data and tell your data story, and (5) test the report to ensure it operates and looks the way you want. In this post will focus on some best practices for the second step in the process: Choose the Best Chart.

To choose the best chart to display your data, there are some best practices (or general rules) you should follow.

Use the simplest graph

There is no need to display something in a word cloud that can be more easily understood in a bar chart. Unusual chart types might muddle your data message and make it more difficult for people who use assistive technology (like screen readers).

Select any image to see a larger version.
Mobile users: To view the images, select the "Full" version at the bottom of the page.

In this example, both the bar chart and the word cloud are displaying the same information (Total Profit by Company). The profit details are easier to interpret and more easily compared in the bar chart versus the word cloud. Word clouds should be avoided if analytical accuracy is desired because it is difficult to compare the relative sizes of different words due to the number of letters in each word and the size of the letters.

Pro Tip! As you find yourself creating report objects and modifying options, you can create object templates with those options applied to speed up development time in the future. To create an object template, right-click the modified object and select Save to Objects pane. The object is added to the Objects pane with a default name (which can be modified) and available for you to use in your reports. For more information about object templates, view SAS Visual Analytics: Use Object Templates for consistency and faster report development.

Use visually appealing, easy to understand objects

The type of objects you choose will depend on the audience of your report. Something that is easy to understand for statisticians might not be easy to understand for the general public.

In this example, a Cluster object (located in the Statistics group) is used in a report for statisticians, who understand the concepts of clustering and can easily interpret the various graphs. For a general audience, however, simpler objects (like key value objects and bar charts) are used and explanatory text (using the Text object) is added to explain the layout of the report and state the conclusions.

Use only the most important data

Only include what is necessary to get your point across. If too much data is displayed, your viewer might miss the point of the graph.

In this example, both reports show the same information. The report on the right has additional details displayed in the key value objects (Transaction Year) and additional roles added to the bar chart (Lattice Row and Lattice Column). None of this information adds any additional context for the report and makes the report seem more cluttered and difficult to interpret than the report on the left.

Pro Tip! Lattice columns and lattice rows in charts might not be apparent to users of screen reader technology. To ensure users are aware of lattices, include a text object that describes the lattices and explains how to interact with them.

Keep graphs simple

Only include what is absolutely necessary in your charts. In general, you should try to remove all chartjunk. Chartjunk refers to all visual elements in charts that are not necessary to comprehend the information presented or distract the viewer from the main point of the chart. This can include tick marks, grid lines, or unnecessary labels.

In this example, both charts show the same information. The chart on the bottom includes a lot of chartjunk: the vertical axis, gridlines, labels and tick marks for the horizontal axis, and the fill. The chart on top minimizes the amount of chartjunk making it much easier to focus on the data values.

Use a zero baseline

A zero baseline is typically assumed by users and can lead to misinterpretation or tend to skew the understanding of the chart if it is not used. If you would like to compare values to some baseline other than zero (perhaps to compare the results for a drug trial to the placebo group), use a needle plot (where a zero baseline is not assumed). If you want to highlight the difference between values that are close together, use a dot plot (where a zero baseline is not assumed).

In this example, both charts show profit by area code. At first glance, the chart on the right seems to indicate that some area codes have negative profit. When taking a closer look, however, you can see this is because the baseline is set to $50,000. The chart on the left is a better representation of this data because it starts the baseline at zero, so total profit values are not misread. To compare profit for each area code to $50,000, add a reference line to the bar chart or use a needle plot instead.

Use two-dimensional charts

Three-dimensional effects can distort the data and cause the viewer to miss out on vital information. Luckily, SAS Visual Analytics only produces two-dimensional charts!

In this example, both charts show revenue and costs by quarter. In the chart on the right, costs for 1st quarter 2015 and 1st quarter 2016 are hidden by the revenue values. In addition, comparing the values is very difficult and depends on the angle at which the chart is viewed. The chart on the left shows all data and makes it easier to compare the values for each quarter.

Pro Tip! For multi-line charts (like the chart on the left), it’s a best practice to rotate attributes for data element styles. This makes it easier for color blind users to read the chart by rotating the type of line drawn for the line chart (one line may be solid, the next may be dashed, and so on). To set this option, at the report-level, set the Data Element Style Rotation option to Rotate all attributes.

Choose colors wisely

Use colors to promote focus in graphs, but don’t overuse colors. Less is more. Choose complementary colors that are pleasing to the eye when used side by side and can help improve readability for users who are color blind. The default palette choices in SAS Visual Analytics cycle through warm and cool colors for this purpose. In addition, ensure your colors have sufficient contrast to foreground and background elements and be aware that some colors might evoke strong emotions. These emotions may differ based on the locality or region of the world, so color choices may change depending on your audience.

In this example, the chart on the right uses too many colors (different colors are used for the title, the axes, the legend background, and the graph background). In additional, the colors used to display the left and right axes do not contrast well with the graph background, which makes the values difficult to read. The chart on the left, on the other hand, uses the default color palette in SAS Visual Analytics, so colors contrast nicely with the graph background and each other, which makes it easier for color blind users to compare values.

Pro Tip! Using SAS Theme Designer, you can create custom themes to customize the color palette used in your reports. Use the Options pane to set the theme at the report-level. For more information about creating custom themes, see Use Report Themes to quickly change your data color palette.

Keep in mind, that the data should be the focus of your report. The data helps advance your story and should be what your audience remembers about your report. As stated by the godfather of data visualization, Edward Tufte, “Above all else, show the data.”