Data Visualization Toolkit: Charts Data Considerations

Charts icon

When we think of visually representing data, charts (or graphs) are typically what first come to mind. To avoid having charts that look like Excel defaults or uninspired, here are some tips for creating charts that accurately represent the data and fuel engagement in data-based action.

Charts Data Considerations

After identifying the main purpose of your chart, the next step in effective data visualization is to choose the best chart to display your data. Several common data types and suggested charts are detailed below. Click Show Example to view example charts and click on the example image to download an Excel template for each chart.

Single Number/Data Point

Pie charts are useful in highlighting a single percentage. Multiple pie charts can allow for easy comparisons across single data points.

Comparison of Two or More Data Points

When comparing two or more data points column or bar charts, slope charts, or dot plots highlight the differences between data points.

Survey Responses

Survey responses can be summarized in a 100% Stacked Bar Chart or Diverging Bar Chart. For branching survey questions, Stephanie Evergreen’s blog has a great example of how to use a nested chart.

Data over Time

To show data that changes over time there are many options including the traditional line chart as well as slope charts, dot plots, and column charts.

Data with Negative Numbers

A deviating bar chart is a simple way to display data that includes negative numbers.

Frequency and Distribution

Histograms are used to show the distribution of data and are good displays for frequency data.

A Large Amount of Data

As noted in the Design Principles, using multiple charts or small multiples—several mini charts in one—works well if you want to display many different categories, trends, relationships, or variables. A stacked bar chart is also an effective chart type when you have a large amount of data.

Qualitative Data

When analyzing and interpreting qualitative data, it’s useful to identify common themes in the data. A heat map is one way to “chart” themes.