Data will often contain intrinsic dimensionality. The most common example is a date. Dates are usually entered as single values, such as 12-May-2005, but if you think about it, there is an intrinsic dimensional hierarchy you might find useful when building visualizations: year > month > day, for example, but also year > quarter > month > day. Include time and you have additional dimensions.

Independent data attributes, or columns, can also be arranged in a hierarchy. A common example is geography (region > country > province > city). The order of the hierarchy can vary depending on your exact needs for a given situation. Do you want to arrange your music collection as artist > year > album > song or year > artist > album > song?

You won't always need to use dimensional hierarchies, but they can be a useful and powerful way to navigate your data.