When Pie Charts Go Pear-Shaped

Pie charts are extremely popular for data visualization, often used to break up a report visually and inject a bit of variety. However, it is also one of the most misused charts in data analysis.

As Stephen Few, Principal of the consultancy Perceptual Edge writes of pie charts in his article: Save the Pies for Dessert, 2007: 

“[A pie chart’s] simple charm is beloved by all but a few, welcomed almost everywhere; familiar and rarely threatening. Of all the graphs that play major roles in the lexicon of quantitative communication, however, the pie chart is by far the least effective. Its colorful voice is often heard, but rarely understood. It mumbles when it talks.”

However, whilst this is certainly a scathing review of a chart loved by many, it is rooted in a fair point. The value of data visualization lies in the ability to quickly understand data without trawling through data tables or reading long-winded analyses. For this, there is only one use case for the pie chart over an alternative: when presenting parts of a whole in a volume of less than three.

Where pie charts go wrong

We’re all pretty familiar with pie charts. Aptly named because they look like their namesake, pie charts take your data and visualize them in variously sized sections around a central point. Each section of the pie shows a part of the whole, so a segment of data that occupies a higher percentage than another will have a larger slice.Pie charts are a popular data visualization tool because of their simplicity of visual design

, but this same simplicity is a double-edged sword.

Despite the simple design, the more segments a pie chart has, the harder it is to read.  For example:

One way to counter this issue is by adding percentage labels to the chart, but one can’t help but wonder – if you’re looking for the biggest number, would an ordered data table get readers to the answer faster?

Where pie charts are Brilliant

Where pie charts excel is when they represent percentages. We'd recommend you only use pie charts to describe parts of the whole and never use them for anything else.

A brilliant pie chart clearly shows the dominant answer to a question with limited options. Examples may include:

-       ‘Yes’, ‘No’, ‘Maybe’

-       ‘Agree’,‘Unsure’, ‘Disagree’

-       ‘Often’ ‘Sometimes’ ‘Never’

EXAMPLE ONE

In this example, with only three answers to choose from, the data respondents give create neat and legible slices for your pie chart.

EXAMPLE TWO

In this even simpler example, we can easily see the ratio of people who drink their coffee with vs without sugar.

However, in both examples, if what you are looking for is the exact percentages of each category, a gauge may serve your report better, or data labels. In the example pictured, this chart is best if understand which category has ‘more’ and which has ‘less’ is the key takeaway.

Alternatives to Pie Charts

What are the best alternatives to a pie chart? If you asked Stephen Few, he’d probably say “anything.” Brilliant Assessments come with various alternatives for presenting ‘parts of a whole’ relationships.

Stacked bar charts are a great contender.

While pie charts are great for simple things, leaving more complicated tasks with the bars is the way to go. Not only do they represent percentages just as well, but they can handle more complex data sets and still be understood at a glance.

Choosing the best chart for the task is crucial to your report's success. Pie charts, while popular, are commonly misused and can cause more confusion than clarity. They are powerful for small options and to present parts of a whole, but for almost everything else, they are best saved for dessert, as Mr. Few says.

Be Brilliant with Assessments

Interested in learning more about charting and feedback assessments in Brilliant Assessments? Watch our demo video here to see how our assessment software can work for you.