Selecting the appropriate chart type is crucial for crafting a clear data visualization.
Making an incorrect choice can lead to misinterpretations, confusion, and hinder the audience’s understanding of your intended message.
With numerous chart types available, the selection process may initially seem daunting, but it becomes much more manageable once you understand the guiding principles.
When deciding which chart to use, you should always start with the most important question:
What are you trying to communicate with your data? What is the story you want to tell?
Do you want to show change over time? A list of rankings? Correlation between two variables? Once you know that, the options are clearer.
The Financial Times’s data visualization team has made a great visual vocabulary for this. We’ve used their categories to help you pick the right Flourish visualization.
Visualizing change over time in data visualization allows us to perceive patterns, trends, and fluctuations.
The time periods can span from minutes to months, years and decades.
For example, if you need to showcase the revenue for your business in the last ten years, a line chart would be the best option. If you want to add projected revenue for future years and want to display uncertainty in the projections, go for a fan chart.
Size comparisons can be relative or absolute and usually show a ‘counted’ number rather than a rate or per cent.
Say you want to compare populations across different countries, a bar chart is the standard way to achieve this – or a lollipop chart if you want to draw more attention to the data value.
Part-to-whole charts show how much of a whole an individual part takes up.
Sometimes, we need to know not just a total, but the components that comprise that total. While other charts like a standard bar chart can be used to compare the values of the components, the following charts put the part-to-whole decomposition at the forefront.
Visualizing election results is a popular use-case for part-to-whole visualizations. For example, the percentage of people that voted for a certain party can be visualized using an arc chart or stacked column/bar chart.
Use correlation visualizations when you want to show the relationship between two or more variables.
Let’s say you have collected data on advertising spend and the corresponding sales revenue for a series of months. Using a scatter plot, you can visualize the relationship between these two variables to understand if there’s a correlation between advertising spend and sales revenue.
The chart types below can be used to plot the variables against one another to observe trends and patterns between them.
Sometimes we are not only interested in the magnitude of the variables but also in the relative ranking of all the variables.
Use the below chart types when an item’s position in an ordered list is the most important thing you want to show.
Go for a bar chart if you want to compare and show which of your products, campaigns or employees performs best. If you have collected this data over time, spanning several weeks, months or years, you can highlight changes in ranking over time by using a bar or line chart race.
Another important use-case for visualizations is to show how the data points’ values are distributed.
With the following chart types you can show values in a dataset and how often they occur. Often it’s useful to see the shape of a distribution.
Say you want to visualize the distribution of customer ages to gain insights into your customer demographic. A population pyramid would be a useful template for this – or a beeswarm if you want to emphasize individual points.
Flow charts convey movement over time. Use these to show volume or intensity of movement between two or more states or conditions.
If your aim is to track product flows or user journeys, a sankey is an effective choice. If you have data on the connections of company employees or product groups, go for a network chart to display these relationships.
Spatial charts can show precise locations or geographical patterns in the data.
But, having geographical data does not mean a map is the best option. It should only be used when geospatial patterns in your data are more important than anything else.
For example, if you want to highlight where you have customers, users or readers, either globally or in region, go for a proportional symbol map if you are displaying total figures. For variation in election results, a choropleth map or equalised cartogram are your best options.
If you want more guidance on choosing the right map type for your data, check out this blog.
Deviation charts show variations from a fixed reference point or baseline. The baseline can be zero, a target, average or median.
They can also be used to show sentiment (positive, neutral or negative).
If you are displaying survey results and your aim is to show the percentage of respondents who agreed or disagreed to a statement, go for a diverging stacked bar. A surplus/ deficit filled line is often used to visualize climate change data.
Always keep in mind that choosing the right chart is crucial. Think about what your data really wants to say, and then explore Flourish’s templates to bring out its potential.
Flourish helps you transform your data into incredible visualizations and meaningful stories.
We can’t wait to see what you’ll create with Flourish – tag us on social media and showcase your visualizations!