One of the most frequent questions when working with data is how to choose the best type of chart for the story you have to tell with your data.
Efficient data visualization will bring your story to the next level and capture the attention of your audience. On the contrary, your best efforts will go down the drain, and you will get the opposite effect.
We are conscious that there are different types of charts to choose among, which can sometimes be a bit of a struggle. This article will give you some little tips to find the best data visualization that suits your project.
Start from the story you want to deliver with your data
Even before thinking about colors and other ways to make your chart eye-catching, take a moment and think about the story you want to provide to your audience.
Just remember, data are just another way to tell stories. There’s a story to tell within every data, but the tools you use have no idea of what it is about. That’s why it’s up to you to bring to life, both visually and contextually, that story.
Have you collected your data to tell a story about trends? Or to make some comparison? Or maybe, to talk about a correlation or connection?
Once you have clear in mind the why of your data, choosing the right type of chart makes the difference between simply showing data and making data storytelling.
Think about the size of your data
When deciding among different types of charts and graphs, you have to ask yourself to think about the size of your data. Sure enough, as you may have already heard, some types of charts are not the first option for large datasets because they would be challenging to interpret and require extra effort from your audience. On the other hand, other types of data visualizations will work perfectly with massive datasets.
Establish the type of data you have collected
Figured out the story to deliver and the size of your data, you just have to establish the type of data you have collected. Contrary to what one might think, data are not all the same. They can be continuous, qualitative, quantitative, or descriptive.
Knowing the type of your data will help you to discard some charts in favor of others.
Different kind of charts and graphs for presenting your data visualization
There are tons of charts you can choose from. List them all would be even more confusing; that’s why to make your selecting process more accessible, we will give you an overview of the most common type of charts by telling you when to choose or avoid them.
Bar Chart vs Column Chart
They are, no doubt, the most common types of charts when talking about data visualization. The persistent dilemma is bar chart vs column chart. Which one to opt for? And, going deeper, what is a bar chart used for?
Let’s start by saying, the bar chart is the horizontal version, while the column the vertical one.
When to use: They are a powerful visualization for comparing the length of the rectangles with categorical data. When labels on the X-axis are too long, the horizontal bar chart’s version is a great option. Bar charts are more comfortable to read for long names since their structure is closer to the western way of reading texts (from left to right).
When to avoid: they are not the best option for comparing trends over time.
We have written an article on our user documentation if you want to know about these two types of charts.
Chart type: Lines
As we have already written in a previous article, line charts are good for comparing trends over time. That means they work best for analyzing continuous data.
When to use: every time you hear the word “over time”, it’s the turn for a line chart. They are also a good option for making forecasts and comparing a significant amount of data all at once.
When to avoid: line charts don’t lend themselves well for making part-to-whole comparisons, managing categorical data, and showing quantities of things.
Chart type: Areas
They are similar to line charts with the difference in areas under the lines filled with colors or shadings. It may not seems a consistent difference, but in reality, it has a relevant effect on how people perceive data within the chart.
When to use: Area charts are good at displaying part-to-whole relationships where one part is vast or changes from very large to very small. They are also suitable for showing how quantities have changed over time. If you need to analyze time values, it is best to opt for an area or stacked chart; instead, you can choose a treemap if there is more than one dimension.
When to avoid: displaying several volatile datasets over time and showing nuanced differences in values.
Chart type: Points
Scatterplots are commonly perceived as challenging-to-understand data visualization, but they are helpful in both scientific and marketing fields.
When to use: Scatterplots are used to display the correlation and clustering in vast datasets or to determine the relationship between the two metrics on-axis. A third metric can be added and it will affect the diameter of the circles.
On the contrary, bubble charts can be an alternative to bar/column charts with the difference they encode the data within bubbles not bars.
When to avoid: scatterplots are useless if the values within the dataset are not correlated. Bubbles sometimes can be perceived much less accurately than bars, especially when metric values are similar to each other.
Chart type: Rings
Excellent at displaying percentages on the total value or proportional quantities. They work better with small datasets. Learn more here about ring charts.
When to use: visualizing the part-to-whole relationship and displaying that one slice of the total is relatively small or large.
When to avoid: with vast datasets, precise comparisons, or when percentages do not sum to 100%.
To sum up, to choose the right type of chart, you need to keep clean in mind these three main factors:
- firstly, the story you want to deliver with your data
- secondly, the size of your data
- and finally, the type of data you have collected
They will make your selection process much easier and avoid the risk of letting you choose the wrong chart. Remember to don’t overload your data visualization with unnecessary elements and keep your chart as clean as possible.
Ready to make your first chart? Start now here