The Four Scales of Measurement of Data: Learning Statistics

The study of statistics is an essential component of everyday life. Data has been classified into four scales of measurement so that it can be easily interpreted universally. The scale is chosen depending on the information that the data is intending to represent. The four scales of measurement of data are nominal, ordinal, interval, and ratio. Each plays a different, yet very important role in the world of statistics.

The nominal data measurement scale is used for data that is expressed with the purpose of identifying some kind of attribute. It can be expressed using either a numeric code or some kind of nonnumeric label. For example, if a university wants to classify its students into groups based on their major, they can express the information by labeling the groups with their respective major names (business, communications, health care, etc). Additionally, they can give the groups a numeric label, for example the number one could represent business, two for communications, and so on.

The ordinal data measurement scale is used when you want to classify information based on a specific order or rank that is necessary. Like nominal data, the information can be expressed either in a numeric or nonnumeric way. For example, if a school wants to classify its students based on the year of schooling that they are in, they could label the groups as freshman, sophomore, junior, etc. Additionally, they could give numeric codes to the groups by classifying ones as freshmen, twos as sophomores, and so on.

The interval data measurement scale is used for numeric data that is expressed in intervals of some kind of fixed measurement. For example, if a school is classifying students based on the results of their ACT scores, then they could say that student A scored a 25 and student B scored a 30. After stating that, they could see that student B scored 5 points higher then student A.

Finally, the ratio data measurement scale is used to express the ratio of some of the values of interval data. For example, is a school is trying to create a data set of how many credit hours students have taken, they could state that student A has taken 10 hours, while student B has taken 20 hours. Using the ratio method, they could say that student B has taken twice as many credit hours of classes then student A took.

Source:

Masterson, John. Lecture: Business Statistics. December 11, 2009