Anything but bars: The 10 best alternatives to bar graphs


What if you want to show a comparison but can’t bring yourself to draw another bar graph?

I don’t ask this question lightly. A bar graph or a line graph can illustrate the vast majority of data; they’re the two workhorses of the dataviz world.

But let’s be honest: bar graph after bar graph gets boring. That’s when you want to have an alternative or two up your sleeve.

Here are the 10 best ways to illustrate comparisons without using a bar graph.

 

Why we use the bar graph so darn much

The most common types of graphs — line graphs, bar graphs, and pie charts — are common for a reason.

Each graph type specializes in telling a certain type of story. Line graphs show trends over time. Pie charts (and their donut brethren) show portions of a whole. And when it comes to comparisons, bar graphs truly shine.

Bar graphs play on humans’ natural ability to judge distances. We’re great at determining which shape is longest, and by how much. The bar graph takes advantage of this ability by replacing numbers with rectangles.

Line graphs, bar graphs, and pie charts are common because they’re the simplest, most effective ways to tell each story type.

 

The scenario: Comparing values for one variable

I sometimes take part in the Data for a Cause Challenge, a project led by Olga Tsubiks that lets data visualizers play with nonprofit and cause-driven data. They recently featured data from UNEP on protected areas. Since I’ve been working with the data lately, I thought I’d use it as today’s example.

The dataset covers hundreds of thousands of protected areas around the globe, each with details on locations, sizes, governance, and more.

There are many ways to show multifaceted data like this, but we’re keeping it simple today. We’ll just compare how much of each North American country is set aside in protected areas. (For ease of calculation, I’m ignoring marine reserves and parks that cross country borders.)

Our bar graph is pretty straightforward: three numbers for three countries. As a percentage of total area, it looks like Mexico has set aside the most land in North America, followed by the United States and Canada.

The bar graph gets the job done, but it could be more exciting. Let’s see what else we can do with the data.

 

1. Row chart

“Hang on,” you might say, “this graph looks awfully familiar.”

And you’d be right. The row chart is basically a bar chart that has been rotated 90 degrees. But it’s not a total cheat; this subtle twist (har har) could come in handy.

If your data labels were a bit longer — say you wanted to spell out “The United States of America,” for example — the vertical bar graph would quickly run out of room. Excel usually switches labels to a vertical or diagonal orientation, but both are hard to read.

The row chart solves this problem by offering more layout flexibility while keeping the bar graph’s rugged good looks and charm (and effectiveness). If you’re writing a report, you can arrange the graph to match the width of a column of text without taking much vertical space.

 

2. Radial column chart

 

The radial column chart is a bar graph that’s been curled around on itself. In the example above, I’ve added Central America to the mix to make things more interesting. (On a related note: nice work, Guatemala!)

Unlike most of the other graphs in this article, I made this radial column chart in Illustrator. As far as I know, there’s no way to make it in Excel.

To be honest, I have mixed feelings on the radial column chart. One the one hand, it’s very eye-catching, and could go nicely with a “hitting our targets” kind of theme. On the other hand, it’s very difficult for a reader to compare values because no two columns are truly side-by-side.

To help address this issue, I’ve added labels and faint concentric rings. Labels tell the reader exact numerical values, while the rings help the reader gauge distances, just like on an Excel-default bar graph.

 

3. Donut chart

 

 

Ah, our old friend the donut chart. I wrote a full article on donut charts a while back, but here’s the gist: a donut chart is essentially a hybrid of pie charts and bar graphs. You read it like a bar graph, but it has the friendly, space-conscious shape of a pie chart.

You’ll notice that I’m using areas here, not percentages. Here’s what the graph would’ve looked like as a row of percentages. This is the WRONG way to represent the data — can you see why?

This donut chart is WRONG – please don’t do this!

It isn’t just that the percentages don’t add up to 100. If you group numbers, people will assume the things they represent belong together.

In the row chart, putting percentages on separate rows made it clear that each number was the percent coverage of a different country. But when I throw those percentages into the same pie chart, they now seem to say that the countries protected 17, 13, and 22% of the same land mass — totally incorrect.

 

4. Stacked row chart

 

If you’re crunched for space, it might be time to put your ducks in a row. The stacked row chart lines up your data so everything is in one relatively narrow line. Think of it as a flattened-out donut chart — which is why the areas are back again instead of the percentages.

As with the donut chart, we need to be careful about what we imply. Stacking a line of percentages would send the same incorrect message as the second donut chart.

 

5. Bubble chart

 

In a bubble chart, the area of each circle reflects the size of your number — in this case, the percent of land protected in each country. I’ve discussed the challenges and potential pitfalls of using circles before, but know that circles are still great if you’re looking for a friendly, eye-catching way to represent data.

I’ve lined up the bubbles in the example above, but you could also cluster them in a circle or other shape depending on the needs of your layout. And don’t forget your labels! When the numerical values are relatively close together (as they are above), they save the day by helping your reader make precise comparisons.

 

6. Icons

 

Circles are fun, but they don’t have much to do with the topic at hand. Here, I’ve replaced our bubbles with trees to represent protected forest cover.

I prefer to use the same icon for all categories to help my reader compare countries directly. If the US had a pinecone and El Salvador had a banana leaf, the variations in each icon might overshadow the different sizes. In short: keep it simple to let your data shine through.

 

7. Dot matrix

 

The dot matrix is great for highlighting parts of a whole. In the example above, we’re using it to illustrate land protection in the US alone. The whole matrix represents all US land, while the orange circles represent the fraction that has been protected.

Like the other graphs in this article, you’ll still need to be careful of how you combine your numbers. If you wanted to compare the three countries, you could put three matrices side-by-side — but clearly separate. If they get too close, you might give your reader the wrong idea.

The dot matrix can also be good for combining multiple categories — as long as it makes sense to do so.

 

In the example above, I recalculated each country’s protected areas, this time as fractions of the entire North American land mass. By using different colors for each country in the same matrix, I can show each country’s contribution to the whole.

 

8. Pictograms

 

 

Pictograms replace the bar chart’s boring rectangles with images; our old friend the forest, in this case. Each percentage point gets a tree in my graph, but you could always scale things differently — a single tree could represent 5% or 100,000km2, depending on the needs of your data.

 

9. Choropleth

A choropleth uses color saturation to represent values; in the example above, it’s the percent of land protected by each country.

Labels are very important for choropleth maps, because your reader can’t just eyeball your map and say “yeah, that shade of orange just screams 17%.” The colors give your reader a hint about the value, but you’ll still need a label to finish the thought.

I made these in Illustrator, but apparently it’s also possible to make them in Excel using an add-in. This tutorial seems to know what’s up — please report back if you give it a shot!

 

10. The big ol’ number

 

 

Last but not least: the big ol’ number. If you only have 1-2 numbers to highlight, there’s nothing wrong with simply writing it out.

Remember: never illustrate for illustration’s sake; only do it if it helps your reader!

Excel, Word, and a host of other online and offline tools can help bring a bit of artistry to your cold, hard data.

 

Which bar graph alternative is your favorite? Which ones did I miss?

Let us know in the comments below!

 


2 Comments

Cancel

  1. Liz

    Hi
    Explained really well 🙂 Love the diagrams.
    I’ve tried to make a chloropleth using a different tutorial in Excel before, it’s pretty handy for Geography.
    Thanks for giving me some ideas for my assignment report

    Liz

    Reply