There is a hierarchy of four levels of measurement: nominal; ordinal; interval; and ratio. Nominal measurement is used for discrete variables with unordered categories. Ordinal measures are also discrete, but it is possible to order the categories. Interval level measurement involves the assignment of a number to each observation, so that the difference between the numbers is theoretically meaningful. For example, the difference between 10°C and 20°C is equivalent to that between 20°C and 30°C. However, with interval level measurement there is no true zero point, so it is not possible to say that 20°C is twice as hot as 10°C. Finally, the ratio level of measurement is the same as the interval level with the addition of a true zero point, so it becomes possible to make statements about the ratios of values. Whilst there is often a natural correspondence between the level of measurement and the nature of the variable, in practice some variables are measured or analysed in different ways. For instance, age is usually measured in years, which is a ratio scale; someone who is 40 is twice as old as someone who is 20. However, sometimes age categories are used so that measurement is at the interval or even ordinal level.
— Stephen Fisher
