In a stacked bar chart, each bar represents a whole set quantitative data. A bar is then separated into segments so that each segment represents a single item of the corresponding set. Stacked bar charts are therefore an alternative representation for quantitative multiset data.

Use the stacked bar chart to display multiple sets of quantitative values. The sets all share the same scaling, i.e. each set has the same number and types of values - the single values of two different sets are directly comparable to each other.

Create a Cartesian coordinate grid and draw the first series of datasets as in a stacked bar chart. Apply the second series of patterns by creating the respective rectangles and “stacking” them on top of their corresponding counterparts from the previous series. Repeat with all subsequent datasets.

Stacked bar charts let the user directly compare the total magnitude of several datasets within one diagram while providing additional information about the composition of these sets from their single items. The value of a whole dataset is the most important information to retrieve from this visualization - the magnitude of single items can be read and interpreted, but their precise evaluation is hampered because the segments of a bar do not share the same base.

The Atlas of Globalization presents a stacked bar chart to display the proportions of globally traded goods from 1970 to 2000.