The different chart types

As we’ve seen, there are seven different chart types to choose from. But which one should you use?

The answer depends entirely upon your data and the story you’re trying to tell. Was your data collected to track performance over time? For which metric? Or was it captured to compare collections? To measure the difference between one value set and another?

Considering why your data was captured and what metric your stakeholders are interested in will make choosing the right chart a much simpler task.

Measuring progress over time

If your stakeholders are concerned with progression over time, i.e.

• Is the company selling more or fewer items than in previous years?
• Are the staff reporting more hazards and injuries or less?
• Are the expenses increasing or decreasing?

Whenever the measure of an item’s value is time, a line chart or an area chart is an excellent choice.

Line chart

If there are multiple collections to be measured individually on the same timeline, a line chart will illustrate the results best. Items within each category are listed along the horizontal axis and values are distributed evenly up the vertical axis. Each data point (timeframe) is connected to the previous data point and the next data point with a line progressing upwards or a line progressing downward in the same colour. Each collection (series) is depicted with a line of a different colour.

As line charts show data over regular time intervals on an evenly scaled axis, they’re ideal for illustrating trends in data at consistent intervals, like months, quarters, or financial years.

Area chart

If there’s only a single series being measured in the timeline, an area chart may be a better choice. An area chart illustrates the volume of each data point, as well as the progression over time, effectively presenting how much each data point is worth, and the increase or decrease over time.

When there are multiple series, a stacked area chart will show the sum of all categories stacked. Instead of illustrating the value of each series item, a stacked area will show the total of all series and depict the relationship of each series item as its portion of the whole.

Compare one-dimensional data

Pie chart

If your stakeholders are interested in comparing the value of just one metric over the collections, like…

• Which product line is the most successful?
• Which work site records the most health & safety incidents?
• Which employee logs the most billable hours?
• Which plant produces the most product?

When measuring a single factor for comparable items against each other, a pie chart will illustrate the value of each category as a percentage of the total of all items in the series.

Comparing collections with 1 or 2 dimensions

Bar or column chart

When there are multiple factors in each collection, i.e.

• How many incidents per risk status have been reported for each worksite?
• What are the operating costs, per expense type, incurred by each department?
• What is the total volume of product produced by the plant per month?

A clustered or stacked bar or column chart will present the results as a comparison between the collections. A stacked column or bar will better illustrate the total per item, broken down into subcategories, whereas a clustered column or bar chart will better illustrate the comparable subcategories per item.

A column chart displays categories along the horizontal axis, secondary categories in the legend, represented by different colour columns, and values along the vertical axis, whilst a bar chart shows categories on the vertical axis, subcategories in the legend, represented by different colour bars, and values in the horizontal axis.

Comparing multi-factorial data series

Surface chart

A surface chart is ideal for uncovering peaks and valleys in data when plotting 2 sets of numeric data on the same chart. Like a heat map, instead of showing categories in the legend, the different colours in the chart are value ranges. Each value band is assigned a different colour to highlight the highs and lows.

Radar chart

Radar charts (or star charts) are used to plot the performance of multifactorial metrics and determine correlations of components that lead to success or failure. Radar charts are usually plotted with many stars on each web and each star represents one data series.

To change the chart type of a chart already created; from the Design tab of the ribbon click ‘Change Chart Type’ to display the dialog box. Select the category and the preferred chart type from the dialog box and click OK.

Share your thoughts

When selecting a chart type, how would you consider the purpose of your data and the story you want to convey? Can you share an example where choosing the right chart type significantly improved the effectiveness of a data presentation? Conversely, can you think of an example where selecting the wrong chart type made a data presentation ineffective or difficult to understand? Reflect and share in the comments below.