Look for differences in the heights of the bars. The bars show the value for the groups. Refer to the scale range of the y-axis to determine the actual differences.
For example, the following bar chart compares the counts of different types of paint flaws. Peels are the most common, followed by Scratches, Smudges, and Other. The counts range from about 6 to 15.
Hold the pointer over a bar to view a tooltip that shows the value of the bar.
Compare bars within the clusters to understand the proportions of subcategories within each main group. Compare bars from the same subcategory across clusters.
For example, the following bar chart shows the mean light output of three different types of glass. Each type of glass was created at three different temperatures. The temperature that has the highest mean for light output is 150 degrees. The glass type that has the highest mean for light output is Glass Type 1.
If you have more than one categorical variable, you can create a different view of the groups by switching the order of the variables in the dialog box. The categorical variable that you enter first in the dialog box is the outermost variable on the axis.
For example, the following bar charts show mean light output for each combination of glass type and temperature. When temperature is the outermost variable, it is easier to see that glass type does not appear to affect light output unless the temperature setting is 150.