There are no hard and fast rules when it comes to which widget is best to visualize your data. This guide gives you a basic outline of which widgets seem to work best for different types of situations. When it comes to visualizing data, don't be afraid to "break the rules" to find what works best for your data.

Deciding Which Data is Most Important

Before considering a type of widget to use, you must decide what is most important in your data to get a point across. Is the most important data to display in the visualization an amount or is it a label?

It isn't always easy to discern if the important data to display is a dimension or a measure. A bar chart can help you decide what should be emphasized.

- To emphasize the height of the bar(s) is to emphasize the measure.

- To emphasize which bar(s) is to emphasize the dimension.

Measure (amount)

An amount or an observation; the values that aren't known beforehand
Raw values and calculated values
Typical examples: number of page visits, time spent per ticket, money earned

Dimension (label)

Labels and axes
Dimensions are the context for your measures, such as where it was taken, when it was taken, what is was taken on
Typical examples: date, country name, URL, category

Choosing a Widget Based on How Data Should Be Displayed

This section on the basics of choosing a widget is based on the work of Dr. Andrew Abela. He describes how to best display visual data in his book Advanced Presentations by Design.

After deciding if the data you want to display is a measure or a dimension, refine your widget search by defining a data visualization category.

Composition

The composition category is good for giving viewers a summary or first look at data. It is especially suited for helping to understand large amounts of data.

Examples of widgets that excel at displaying overall data include: pie charts, sunburst, percent stacked bar chat, area, and stream charts.

Comparison

The comparison category is good for giving viewers a clear way of understanding what data is bigger or smaller in the dataset. These are more useful after the dataset has been narrowed down to analyze a small set of data points.

Examples of widgets that excel in displaying the measure of data points include: bar charts, polar area charts, and line charts.

Relation

The relation category is good for viewers to understand how different data points in your dataset interact with each other. The goal isn't to show a particular measure but how different dimensions compare.

Examples of widgets that excel in displaying how data points compare with other data points include: map graph, circular network, scatter plots, and network graphs.

Distribution

The distribution category is good for viewers to understand the overall shape of the data. When the importance of the visual isn't a specific point but coronations and trending within the data.

Examples of widgets that excel in displaying clusters and trends include: scatter plots, histograms, and line charts.

Browser not supported