Levels of Measurement

[Levels of Measurement] [Descriptive Statistics for various levels] [Tools of Inference]

The Level of measurement of a variable is a critical factor in determining what kinds of tools may be used to describe the variable, and what means of analysis may be used for inference about the variable. In short, the level of measurement determines or constrains the types of descriptive and inferential statistics that may be applied to the variable.

Levels of Measurement

In addition, there are two other categories that often get used

Dichotomous variables are variables that only have two values. This level of measurement may be treated as nominal, however, sometimes an ordinal quality may exist.

Some examples

Nominal Variables

Nominal variables are those which can be named, but not quantified. Examples include

Nominal variables may be coded with numbers, but the magnitude of the number assigned is arbitrary. Changing the coding scheme will not change the inference.

Ordinal Variables

With ordinal variables, there is a rough quantitative sense to their measurement, but the differences between scores are not necessarily equal. They are thus in order, but not fixed.

Interval Variables

Variables or measurements wghere the difference between values is measured by a fixed scale.

ratio variables on the other hand are at the other end of the scale. ratio variables are numbers with some base value. percentages are perhaps the best indicator here.

Descriptive Statistics for Various Levels

Tools of Inference