2. Statistical Variable Types

In statistics we collect variables into several types. These types tell us what types of mathematical properties are allowed.

2.1. Categorical or Binned

The most basic variable type is a label or a categorical type. This can be a list of cities, class numbers, or sub-particle components. For categorical data, basic arithmetic has no meaning.

2.2. Ranking or Ordered

These are similar to categorical data, but they can be ranked. This would be like letter grades. Now you can say one value is greater or less than another value, but other mathematical functions, such as addition, still don’t have meaning.

2.3. Interval or Continuous

For continuous interval data, addition now has meaning, but there is no defined zero so multiplication does not. For example, Celsius and Fahrenheit are interval variables. 20 degrees Celsius is not twice as hot as 10 degrees Celsius but 10 degrees Celsius is half way between 0 and 20 degrees.

2.4. Ratio or Continuous

These are the typical physics and mathematical variable. Continuous ratio variables now can be multiplied. For example, in Kelvin, twice as hot now how meaning.

2.5. Suggested Reading

• Table 1.3 in Wall & Jenkins