When NOT to Use a Line Chart

Original Line graph showing the Gender Development Index for eight South Asian Countries in 2017 and horizontal bar graph of same data.

Line charts are a classic tool for visualizing trends over time (e.g., days, weeks, months, years etc…). They are most effective for:

Showing trends in a variable over time

Line chart showing fictitious data about organizational productivity.
Data source: Generated data (fake).

Showing trends in a variable over time for multiple categories (if the variable is measured the same way for all categories)

 

Line chart showing participation in adult education in Europe.
Data source: https://unesdoc.unesco.org/ark:/48223/pf0000265866.

Or showing trends in different variables over time (if the variables are measured on the same scale)

Line chart showing rates of unemployment for adults with less than a high school education; a high school education and some postsecondary educaion.
Data source: https://familywelfare.umaryland.edu/presentations/wa2011.pdf.

You could even use a slope chart (a line chart that only shows two points) to visualize differences between two dates:

Slope graph showing changes in nonprofit sector spending per resident in select US states (Pennsylvania, Maryland, Virginia, Delaware, and New Jersey).
Data source: https://delawarenonprofit.org/wp-content/uploads/2017/09/state-of-the-sector-summary-8.25.17.pdf?x93731.

or differences between two categories on multiple variables:

Slope graph showing differences in employee feedback between an overall company and a Team at the company.
Data source: http://www.storytellingwithdata.com/blog/2014/03/more-on-slopegraphs.

And while there are few hard-and-fast rules when it comes to data visualization, one thing is for certain: line charts are not suitable for comparing multiple categories at one point in time for a single variable.

Let me show you what I mean.

Last week, I was perusing the United Nations Development Programme website, and I came across the following chart in one of their country reports:

Line graph showing the Gender Development Index for eight South Asian Countries in 2017.
Data source: https://hdr.undp.org/system/files/documents//nhdr-2019iipdf.pdf (page 44).

The graphic is a line chart showing the Gender Development Index for eight South Asian Countries in 2017. The Gender Development Index is a measure of gender disparities in human development achievements along three dimensions: (1) Long and healthy life; (2) Knowledge; and (3) Standard of living. The scale ranges from 0 to 1, where 1 indicates perfect gender equality and a value closer to 0 indicates near perfect gender inequality.

A line chart in this instance is not appropriate because the data do not show trends over a period of time. Rather, the data presented are from the same point in time (i.e., 2017) for different South Asian countries.

So, what should you do instead? One alternative that packs a punch is the classic (horizontal) bar chart.

 

Redesign: the horizontal stacked bar chart

An easy choice is the (horizontal) stacked bar chart. It is clean, effective, and easy to interpret:

Redesign: Horizontal bar chart showing the Gender Development Index for eight South Asian Countries in 2017.

Add an informative title and highlight the country of interest (which in this case was Bhutan) and you’re ready to go:

Redesign: Horizontal bar chart showing the Gender Development Index for eight South Asian Countries in 2017, with Bhutan emphasized.

So, next time you’re about to use a line graph, make sure your data falls into one of these categories:

  • Captures trends in a variable over time

  • Captures trends in a variable over time for multiple categories (where the variable is measured the same way for each category); or

  • Captures trends in different variables over time (where variables are measured on the same scale).

Previous
Previous

Sometimes it Pays to Swap Axes

Next
Next

Think About (the Amount of) Ink