Clustered Bar Graphs

A clustered bar graph is a type of bar graph that allows for the display of two categorical variables. Like the typical bar graph it compares data from a sample that pertains to specific categories or attributes of a categorical variable to another property, usually a frequency, about those given attributes. The clustered effect allows two categorical variables of a sample to be graphed at once, with each common attribute being grouped together. This type of graph allows for:

  • Easy comparison between components of a sample
  • Allows for quick generalizations and conclusions to be drawn from  data

Different bars should be given different colors so the bars of each categorical variable can be distinguished from one another.

A good example would be the graph to the right. ClusterGroupBar_SAS92M3It compares the frequency of different types of vehicles distributed on three continents. The categorical variables are the origin (continent) and type of vehicle.

A brief definition of bar graphs, as well as several types of bar graphs, can be found here. This site has an example of how to make a clustered bar chart using SPSS  (Note: the vertical axis here is made incorrectly). Also, Forbes has a good article detailing the differences between a histogram and bar chart.

As a bad example, this graph compares the gas mileage of several types of vehicles on three different continents. Here the explanatory variable is the type of vehicle, however, the response variables are both the continents and gas mileage. This is a common mistake, and can be seen on many clustered bar charts if you were to do a Google image search.


Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s