Keeping it Simple | Data Presentation & Bar Graph Readability

Keeping It Simple | Data Presentation & Bar Graph Readability

Before reading this text, I’m guessing you’ve already peeked at the bar graph below.  It’s human nature for our eyes to be drawn towards a visual – that’s why they’re great for depicting data!  Take a fresh look at this graph, but this time be cognizant of where your eyes go to process the details.

Bar graph BP-1.png

Our natural inclination is to start processing visuals in the same way we process text – starting at the top left and zigzagging across and down the page in Z patterns.  Typically, we first take note of the graph title, then back to the upper left to look at the largest values of the bar chart, move across the tops of the bars, then make our way to the bottom left for the category titles.  At that point, you may need to circle back to the upper left again to tie the category titles back to the values.

This is all done in a matter of a few seconds, but can it be simplified?  Plus, when working with visuals, it’s beneficial to look at alternatives to find the best one.  Is there another way to illustrate this data that’s easier and/or quicker to analyze?  Look below at the same data represented in another way.

Bar graph BP-2.png

If we read through this visual in our standard Z formation, we can quickly glean – by category title – which visit types are most prevalent.  But, there are more advantages to using a horizontal bar graph instead of a vertical bar graph.  The former allows the category titles to be read much easier, especially when they’re lengthy – you don’t need to cock your head to the side like a curious dog to read the diagonal labels.  Also, compared to the vertical bars, the actual bars are longer on the horizontal version (go ahead and use your finger to measure, I’ll wait).  This allows for better differentiation between how much bigger some values are in relation to the others.  Yet another advantage is the ease at which your eyes can compare bar lengths by scanning up and down as opposed to left to right.

However, horizontal bar graphs aren’t always the best choice for your data presentation.  One case where vertical bar charts are best is when graphing data over time.  Dates are best understood when the x-axis is treated as a timeline, so turning that graph on its side would be confusing. Another case to use vertical bar charts are ordinal, or sequential, data series.  For the same reason as a timeline, if looking at a count of salaries or ages, the data series should be put in order from smallest to largest on the x-axis. 

Another way to think about which orientation to choose is to consider how the bars are being sorted.  If the data are sorted by the size of the bar, then either type of graph will work.  Aggregating data by category as shown above allows flexibility to try horizontal or vertical bar graphs.  But, if the data are sorted by the labels – as they would be with ordinal data – then the vertical bar graph makes the most sense. 

If you are looking to display data using bars, it’s common to use the vertical bar chart.  There are scenarios where that’s a desirable choice, but be aware of the horizontal bar chart as an alternative.  Among other factors, consider label lengths and data series types when making your choice. But, the best way to determine which option to use is to compare them side by side and pick the one that’s easiest to interpret.

Bonus tip: consider eliminating clutter by removing gridlines in graphs where data labels are used.

Bar graph BP-3.png

Removing the gridlines makes it easier to read the data labels which improves the overall bar graph readability. It also improves the appearance of your graph.

Sometimes bar graphs aren't exactly what you need either. Check out our blog post Pie Charts Hurt Your Outcomes Data | Pros and Cons of Pie Charts to see when you should use pie charts to represent your behavioral health data. 

Don't be afraid to try your graph both with and without data labels to see which you like best!

New Call-to-action

Topics: Business Process, Industry Insights, Things You Should Measure