Types of Attributes

Knowledge about the types of the attributes can provide guidance in selecting graphical encoding strategies.

For any type of dataset, there are 3 types of attributes:

• Categorical: two or more categories but no intrinsic ordering, ex. hair color: blonde, brunette, black

• Ordinal: two or more categories with intrinsic ordering, but distance between values is unknown (therefore, no arithmetic operations can be performed on the values), ex. economic status: low, medium, high

• Quantitative: the values represent a measured quantity, and the distance between them is known/can be computed (therefore, arithmetic operations can be performed on the values), ex. height, weight

Note: Just because the values are numeric, it doesn't mean that the attribute is quantitative. For example, 'order id' will have numeric values, but it is ordinal (or even categorical, if the ordering is unknown), but not quantitative, because the values do not represent a measured quantity.