[This is a guest post by Severino Ribecca*, as part of a series dedicated to each individual kind of chart that he has read into as part of his main research project.]
In this post I’ll be covering the well-know and widely used Bar Graph, a chart that uses rectangular bars with lengths proportional to the values that they represent. Typically, Bar Graphs are used to compare categorical values, with one axis of the chart used to show the specific categories being compared and the other axis is used as a quantitive scale.
Sometimes people refer to Bar Graphs as Column Graphs instead. However, the only difference is in the orientation of the rectangular bars: Bar Graphs display horizontal bars and Column Graphs display vertical bars.
The type of data used for Bar Graphs is discreet, countable data. Alternatively, when the data is continuous, as opposed to discreet, it then becomes a Histogram, a visualisation method used to show frequency or data over intervals.
Displaying negative values on Bar Graphs works well, especially when the bars are vertical like the in the graph below. Having the columns going down convey a better downward direction of values.
Like on the previous post on Area Graphs, the first appearance of a Bar Graph was in William Playfair’s 1786 book The Commercial and Political Atlas. Inside this book, a chart on the “Exports and Imports of Scotland to and from different parts for one Year from Christmas 1780 to Christmas 1781” used rectangle bars to display the amount of goods that were exchanged between Scotland and each country.
What to Watch Out For
One major flaw with Bar Charts is that labelling becomes problematic when there are a large amount of bars. Labelling can also be troublesome on vertical/column graphs, where the label names are long. Therefore, horizontal bars are better for displaying data with long category names.
The Rest Of The Family
Other relations to the Simple Bar Graph include:
• Multi-set Bar Graphs – Multiple bars grouped together to compare multiple sets of data.
• Stacked Bar Graphs – Bars stacked together in segments to compare multiple sets of data.
• 100% Bar Graphs – A variation of a Stacked Bar Graph that displays a part-to-the-whole relationship
• Radial Bar Graphs – A Bar Graph drawn on a polar grid. Length of each bar is determined by an arc-length.
• Radial Column Graphs – A Column Graph drawn on a polar grid. Uses each ‘ring’ on the grid as a scale.
• Population Pyramids – Two Bar Graphs displayed side-by-side to display a population’s age distribution between men & women.
There are also other charts that use the length of rectangles to convey meaning. Span Charts are one, as are Candlestick Charts and Gantt Charts are another. I will try to cover these charts in more depth in the future.
Why Are Bar Graphs So Effective?
So why are Bar Graph so effective for comparing values? A lot of it comes down to how Bar Graphs visually encode data. Visual encoding is when visual information, such a shapes/forms are used to represent abstract information such as numbers. Some charts use angles, others use area and different shades of colours can be used as well.
With human perception, line lengths and positioning in a 2D plane is a more effective way to visually display data for comparison. Ours eyes can more accurately see changes in positioning then any of the other form of visual encoding. Bar Graphs use the positioning of a bar’s end point to encode values.
I’m curious as to why our vision works in the way it does. While I haven’t yet found any explanation in neuro or cognitive science, Steven Few in his article: Tapping the Power of Visual Perception gives an explanation based on evolutionary psychology:
“Our ability to perceive differences in 2-D areas hasn’t evolved to the same level of accuracy as our perception of differences in 2-D position, perhaps because it was more important for survival that our ancestors could detect the exact location of the saber-toothed tiger, rather than its exact size.”
Since Bar Graph are so commonly used, finding software that can generate them is extremely easy. Even if you don’t have any software, web tools like Google Docs and Google Drive have the option to create spreadsheets that can generate charts. Additionally, you can find a list of data visualisation web tools and source code for generating Bar Graph on my reference page.
In the next post I will be looking at Box & Whisker Plots.
*Severino Ribecca is a British graphic and information designer interested in data visualization. Currently he’s building an online library of different information visualization methods called The Data Visualisation Catalogue. You can follow the project’s updates on Twitter (@dataviz_catalog) and support further developments on the Patreon Page.