80 types of charts for data visualization. Part 1: Comparison Charts
by Ivan Kilin — Data Visualization Specialist & Dieuwertje van Dijk — Data Visualization Designer @
Recently we’ve rewritten our essential article about 80 types of charts. Our main goal was to make the information structured and helpful right away. We talk about every chart starting from the look and feel and then going into practical advice on when you should and when you shouldn’t use this particular chart type. We will start publishing parts of this article here. One chart group per each article.
In the first part, we’ll talk about comparison charts. Comparison charts are usually the first choice for data visualization. This is because in most cases we visualize data to compare. Comparison charts are designed to show how values compare to each other. In most cases, a bar chart is the first and most logical choice.
Let’s start exploring comparison charts one by one. Keep on reading!
Bar chart
Alternative name: Bar graph
What is a bar chart?
One of the most common chart types out there. A bar chart is a set of rectangles with a length proportional to the values it represents. Each rectangle — the bar, is a representation of one category.
When to use bar charts?
Use bar charts when you need to compare values that correspond to categories or discrete values. Bar charts work best when visualized values are in the same order of magnitude.
Bar charts are universal. It can be used everywhere, from a school project to a scientific paper. It is widely recognisable. Bar charts are commonly used due to their simplicity. Viewers need just to decode their bars’ length and position, making bar charts very easy to understand. The general public is fairly capable of reading bar charts, so no extra dataviz expertise is necessary. The differences in bar length are easier to perceive, than, for example, differences in size and color. For this reason, bar charts are doing their job well. You should consider using bar charts in your data visualization if the data structure and the actual message allow for it.
When you shouldn’t use a bar chart
The bar chart has a few limitations. First, the axis along bar charts should always start from zero. It makes it impossible to use it for some specific data range that does not include zero. Also, if your values vary greatly in size, it’s better to use charts that use area for representing a value. Alternative options here are an icon chart or treemap. If there are a lot of bars (above 30), it becomes really hard to read the chart.
If you want to improve your dataviz skills and design the best bar chart you can check our bar chart resource page and discover even more pro design tips.
Column chart
Alternative names: Column graph, Vertical bar chart
What is a column chart?
Long story short, you can say that a column chart is the same thing as a bar chart, turned by 90 degrees. Indeed, a column chart is a type of chart that resembles a bar graph with bars positioned vertically.
When should you use a column chart?
The main difference between a column chart and a bar chart is in the usage of categorical labels. Long labels don’t suit column charts because of easy overlapping. But it might be useful if the labels are short and don’t take up a lot of horizontal space.
Column charts can sometimes be useful when showing date-over-time data. But there is a limitation. It only works if there are only a few timestamps. For an extended number of timestamps, it’s better to use line charts.
Column charts are as universal as bar charts and can be used in every domain where charts can be placed. Column charts are better than bar charts for cases when categories have a natural order, like income levels or education levels. Column charts are very easy to read and they are easily recognised by the general public.
When you shouldn’t use a column chart
If the names of the categories are too long, column chart becomes really cluttered and hard to read. Changing the angle of the labels won’t solve the problem, it makes labels unreadable. So if the labels are too long, use bar chart.
You can use our bar chart resource page to learn how to improve the readability of your column chart as well.
Grouped bar/column chart
Alternative names: Paired bar/column chart, Clustered bar/column chart
What is a grouped bar chart?
A grouped bar chart is a multi-series variation of a bar chart where every category is represented by several columns. Each bar is communicating different aspects of the main category. Columns of each category are separated from the other categories using spacing.
When should you use a grouped bar chart?
This type of chart is used to compare multiple bars based on two variables, either categorical or discrete. For example, product sales grouped by product categories. Usually these types of charts are used in reports or presentations. If you need to make a comparison of values that are related to multiple categories and they should be on the same line, use a grouped bar chart.
When you shouldn’t use grouped bar chart
Despite this chart being a variation of a very useful and popular bar chart, it comes with a few guidelines to keep in mind. Placing too many categories makes the chart cluttered and hard to read. Also, there are a lot of cases when poor color choice makes grouped bar charts unreadable.
Lollipop chart
Alternative name: Lollipop plot
What is a lollipop chart?
This type of chart consists of a line and ends with a dot, or a circle. The length of the line represents the value. The dot or a circle used as an extra highlight of the value.
When should I use a lollipop chart?
You can use a lollipop chart when you need to show the relationship between a numeric and a categorical variable.
A lollipop chart can be a sweet alternative to a regular bar chart if you are dealing with a lot of categories and want to make optimal use of space. All the rules that are applied for bar and column charts can also be applied here.
When you shouldn’t use a lollipop chart
When you have enough space for a bar chart, it is better to use a bar chart. Also, when you have values that are very close to each other, bar charts are better. They are easier and more accurate to read than a lollipop because of the data mark that can be hard to compare, especially when the data mark is big.
Bullet chart
Alternative name: Bullet graph
What is a bullet chart?
A bullet chart is a type of chart designed to benchmark against a target value and ranges. Bullet charts resemble a combination of bar/column charts and progress bars. The results are shown in a single bar or column. The ranges bar is constructed based on values from a category that comparison will be based on (for example competitor sales figures). All these values are then divided into a certain number of sub-ranges (in most cases it’s quartiles). Target shows the value which is aimed for. And the bar shows the actual figures.
When should you use a bullet chart?
Bullet chart is a space-efficient chart used primarily for displaying performance data. Bullet chart is a primary choice for cases where multiple specific values should be shown in a compact way. Actual value, target values, reference ranges — all this can be placed in a space of a single bar.
When you shouldn’t use a bullet chart
Bullet charts have limited use. Being a very specific performance or benchmarking chart, is not useful for any other cases.
Dot plot
Alternative name: Dot chart
What is a dot plot?
Dot plot consists of one or more dots related to each category.
When should you use a dot plot?
Dot plots are a good choice for comparison. Especially if you need to place multiple values under one category. It allows you to have a lot of information on the chart and be clean and clear at the same time. Also, it’s important that it can be used for ranges of values. Unlike bar charts, you don’t need the X axis to start from zero. It gives additional freedom for changing the chart range. It allows you to zoom into a certain part of the dataset.
Dot plot packs a lot of information in a small space. A dot plot with only one value per category makes a comparison between those categories very easy. When the dot plot has multiple values per category, you can also compare within the categories.
When you shouldn’t use a dot plot
This chart may need gridlines that turn a dot plot into a chart with a proper context. Visually it might lack some “base”, compared to bar charts.
More about dot plot
We wrote a very interesting article about dot plots.
Make sure to also check our dot plot resource page and discover pro design tips.
Dumbbell
Alternative names: Dumbbell plot, Dumbbell chart, Connected dot plot, Dumbbell dot plot, DNA chart, Barbell chart
What is a dumbbell chart?
A dumbbell is a variation of dot plot with two connected values per category. A dumbbell consists of dots (or circles) and connectors (or lines). Not adding marks and only leaving the connector makes it a range chart.
When should you use a dumbbell chart?
Use it when you want to emphasize the delta (change) between the two values (data points, i.e. two points in time). Also it can work for comparing values in the same category and throughout all the categories of the chart.
A dumbbell chart is a good choice for visualizing paired values. For example, gender pay gap.
When you shouldn’t use a dumbbell chart
Dumbbell charts work for very specific cases of paired values per category. If it’s not the case, this chart makes little sense.
Pictogram
Alternative names: Pictorial chart, Proportional unit chart, Picture graph
What is a pictogram chart?
A pictogram chart is a type of chart that uses icons or symbols, or even small images, to represent data. Each of these icons can correspond to a certain category. Another option is that the icon represents a certain value or a visual representation of an item. Pictogram charts to some extent resemble bar charts, but instead of using a bar, they show icons.
When should you use a pictogram chart?
Some data visualization experts might argue this type of chart is very basic, to the point that it’s widely used in schools and kindergarten. But it’s important to keep in mind that using a pictogram chart helps overcome language barriers and it’s easy to interpret.
The simplicity of the pictogram chart allows it to use it in cases when the audience is not prepared for charts. IKEA is using only images in their assembly diagrams which makes them universal for all languages. It’s not a dataviz example, but it shows how icons can be a powerful tool for multi-language communication.
When you shouldn’t use a pictogram chart
If the audience of the chart are professionals that constantly work with data, using pictogram charts might be an oversimplification.
Icon chart
Alternative name: Proportional area chart
What is an icon chart?
The icon chart consists of icons of a different area. Each icon corresponds to one category. For most cases the placement of icons can be customized.
When should you use an icon chart?
Icon charts are an effective alternative to bar charts when there’s a significant difference between the values. The icon chart uses area rather than length to visualize values, which allows it to display a larger range of values in a compact way.
When you shouldn’t use an icon chart
Icon charts can be tricky to read. Usually readers are not used to perceiving area in correspondence with the value. It can be hard to see the difference if the values have minor differences. Keep that in mind, if you’re planning to use an icon chart in your visualization.
Range plot
Alternative name: Range chart
What is a range plot?
The range plot looks similar to a bar chart. It also consists of rectangular bars. The difference is that a range plot shows two values of a category, instead of just one. These two points are connected with a line or a bar between them. This line indicates the difference, or a gap, between these points and suggests a direction of such change.
When should you use a range plot?
Use a range plot if you want to highlight the difference between two values, rather than showing these values as something separate. A use case example is any sort of demographic gap, i.e. gender pay gap. Compared to the dumbbell chart, the range chart is more focused on the range itself rather than the individual data points.
When you shouldn’t use a range plot
When you want to focus on a certain value related to a category, better to choose another chart, like bar chart or icon chart.
Radial bar chart
Alternative name: Circular bar chart
What is a radial bar chart?
A radial bar chart is a variation of a regular bar chart with the main difference being the circular shape of the chart. The chart itself is plotted on what is called a polar coordinate system. It means that each bar appears in a circle. The larger the value, the longer the bar.
When should you use a radial bar chart?
Try radial chart if you really want to grasp the attention and you want to try something fancy. Also, if the accuracy of data presentation is not your first priority, it is fine to use it.
When you shouldn’t use a radial bar chart
I was always puzzled by this chart and inability to show the values in a logical way. So let me know if you find a legitimate use case for this chart. It is really hard to read and interpret. Inner and outer bars can seem totally different in size and visualise the same value. Inner bars tend to be shorter than the outer ones.
Parallel coordinates
Alternative names: Parallel plot, Parallel coordinates plot
What is a parallel coordinates plot?
The parallel coordinates chart resembles a line chart. Instead of time values, categories are plotted on the horizontal axis.
When should you use a parallel coordinates plot?
In case you need to plot a multitude of categories/dimensions without compromising the readability in a simple 2d space. Every dimension can be presented with a separate axis and each of them can have its own range.
When you shouldn’t use a parallel coordinates plot
The relative simplicity of the chart, yet, adds some limitations. If each dimension has its own scale, just two neighboring dimensions relationships can be followed at a time. You might look at it as a chain of slope charts. Ordering plays a crucial role in this chart. This chart is not an easy one to read, so you need to consider if the data can be presented in some other way.
Radar chart
Alternative names: Spider chart, Spider graph, Web chart, Spider web chart, Star chart, Star plot, Cobweb chart, Irregular polygon, Kiviat diagram
What is a radar chart?
A Radar chart consists of several axes, all coming from the same point in the center (which resembles a spider web). Every axis is connected to a specific attribute of the compared items (the minimum number of attributes is three).
When should you use a radar chart?
Radar charts are used for making a visual “profile”. Each axis corresponds to an attribute. Usually radar charts are used for side-by-side comparison on the same chart or using separate charts for each profile. They useful for getting an approximate picture.
When you shouldn’t use a radar chart
Although it’s a very interesting chart to use, it’s important to keep in mind that it is hard to read. As it is designed in a circular fashion, it requires extra visual perception, in contrast to the more common linear types of charts and graphs. It is often easier to replace it with another type of chart. If all axes in your chart have the same scale, then a bar chart or sometimes a lollipop will suffice. If the axes have a different scale, it’s better to use parallel coordinates.
Nightingale chart
Alternative names: Nightingale’s graph, Nightingale rose chart, Rose diagram, Coxcomb chart, Polar area chart
What is a Nightingale chart?
A Nightingale chart is a variation of a pie chart. It also consists of sectors, but with different radiuses. In this chart, the value corresponds to the area of the sector, not the radius.
When should you use a Nightingale chart?
A Nightingale chart is visually similar to the pie chart, but it has a different function. It’s used mostly for comparison.
When you shouldn’t use a Nightingale chart
In most cases, a Nightingale chart is not the best primary chart choice. It’s really hard to compare the sector areas of a Nightingale chart precisely. It also inherits all the issues of the pie chart. The most obvious is that it’s hard to distinguish sectors with similar values. For these types of cases, it’s better to use bar charts.
Waterfall chart
Alternative names: Flying bricks chart, Mario chart, Bridge chart, Cascade chart
What is a waterfall chart?
In the waterfall chart, the first value (column) typically starts from the baseline of zero, as does the ending value. They are connected by a number of seemingly floating shorter bars that show the intermediate changes. The overall shape of the chart resembles a waterfall.
When should you use a waterfall chart?
A waterfall chart is a type of graph that usually shows positive and negative values of change between two points, which helps in understanding the cumulative effect of these changes (so the net change).
This chart does not only look at the starting value and the ending value of your data set but also shows each individual positive or negative change that happened. As you can imagine, this type of chart is quite useful in financial sectors or human resources, but also in other industries (think of inventories, revenue tracking, etc.).
When you shouldn’t use a waterfall chart
A waterfall chart is not useful if you want to visualize and communicate the absolute values of each parameter, not the influence of this parameter on the whole.
Matrix chart
Alternative name: Matrix diagram
What is a matrix chart?
A matrix chart is a grid of two or more categorical variables. On the intersection of the categories one or more icons are placed, showing the type of the relationship between categories.
When should you use a matrix chart?
A matrix chart is a chart that helps in visualizing the relationship between two or more categorical items. Specifically, it shows the presence and strengths of such relationships. It can have different forms depending on how many groups must be compared.
When you shouldn’t use a matrix chart
You should be careful if you need to display more than two categorical dimensions, as adding more can quickly reduce readability.
Small multiples
Alternative name: Trellis chart, Lattice chart, Panel chart
What is a small multiples chart?
Unlike all the other graphs in this article, small multiples are more of a visualization concept than a graph itself. This is because small multiples repeat the same type of chart in a grid to show different slices of the dataset.
When should you use small multiples?
The main advantage of using small multiples is the possibility of showing three or (usually) more variables presenting different values in the same graph without confusing your audience. The main way of using small multiples is to split charts that are too complicated or cluttered into many simple charts. For example making a small multiple chart out of a spaghetti chart (line chart with too many lines. Another example is making small multiples out of stacked or grouped bar charts.
When you shouldn’t use small multiples
Small multiples work best when there is enough space to show each chart clearly. If space is too limited, lose clarity. Also, if a single chart can show the same story, they are not needed.
Word cloud
Alternative name: Tag cloud, word collage, wordle
What is a word cloud?
A word cloud is a visual cluster of words that vary in size based on how often they appear in the dataset. In other words, the more frequently a word or keyword appears in the text, the larger and often bolder it will look in the cloud.
When should you use a word cloud?
Word clouds can be helpful for analyzing data related to language and narrative studies. It allows you to get a fast approximate overview of the word patterns in a text.
When you shouldn’t use a word cloud
Word clouds only give a rough idea on the frequency of the words within the dataset. The exact area of the word is not always corresponding to the value because it’s really hard to calculate the area precisely. Words have a complex form and it also depends on the font that is used.
Slope chart
Alternative name: Slopegraph
What is a slope chart?
A slope chart consists of two vertical lines representing start and end points. Data points are placed on each of these lines representing values for one category. Two data points are connected with a line, showing the change using the slope of the line.
When should you use a slope chart?
A slope chart can be used as a replacement for a line chart, grouped or stacked bar chart. If there are only two values to compare, it works perfectly.
When you shouldn’t use a slope chart
If there are more than two time or measuring points, the slope chart won’t work. In this case more conventional chart types like variations of bar chart or line chart should be used.
Table chart
What is a table chart?
Table chart is a usual table. It includes the usual attributes, like rows and columns of data. The difference is that the table chart has some additional visual elements that allow it to highlight certain cells. The most popular version of table charts is a table with some condition formatting. It can look almost like a heatmap, but the difference is that every column(s) or row(s) can have a different rule of coloring.
When should you use a table chart?
Table charts are a good starting point for exploiting the data. Similar to scatter plots and heatmaps, it gives an overview of the data. Sometimes it can also work as a final presentation visualization.
When you shouldn’t use a table chart
The possibilities of using different rules for different columns and rows gives not only the freedom of exploration but also might cause confusion for the reader. Table charts are usually not self-explanatory and need some additional notes and guides on where to look at and pay attention to.
Categorical scatter plot
What is a categorical scatter plot?
A categorical scatter plot differs from a regular scatter plot by the presence of a categorical axis. It can be just one categorical axis or both of them.
When should you use a categorical scatter plot?
Categorical scatter plots can be useful for visualizing a set of 2 categorical variables and one numerical variable. An example is visualizing the results of polls. A bubble positioned at the intersection of two categories shows the numerical value that is related to that specific category pair.
When you shouldn’t use a categorical scatter plot
Categorical scatter plots can have limited utility if the number of categories along each axis exceeds 20–30. In this case the representation of the values can be perceived only approximately. At that point, the scatter plot doesn’t add much value compared to just using a regular table.
Quadrant chart
Alternative names: matrix diagram, matrix chart, 4-quadrant matrix chart
What is a quadrant chart?
A quadrant chart is very similar to a scatter plot but it’s divided into equal parts (quadrants) in a matrix view. It can be 2x2, 3x3, 4x4 or other ratios, depending on the aim of the visualization.
When should you use a quadrant chart?
It is useful if you want to group distinct data marks for some specific type of analysis. One of the best and most well-known examples of using the quadrant chart is for a SWOT analysis. The quadrant chart is useful for benchmarking if there are two dimensions that should be taken into account.
When you shouldn’t use a quadrant chart
In case the splitting into quadrants doesn’t make sense or looks forceful.
That’s it for comparison charts. In the next part we’ll explore correlation charts.
You can also check all other parts here:
Part 2: Correlation charts
Part 3: Part-to-whole charts
Part 4: Data-over-time charts
Part 5: Distribution charts
Part 6: Geospatial and other charts
