In: Statistics and Probability
Using data from such publications as the Statistical Abstract of the United States, Forbes, or any news source, give examples of variables measured with nominal, ordinal, interval, and ratio scales.
Nominal-level:
In the statistical abstract for “State Rankings – White Population Alone,” taken from the U.S. Census Bureau, the states are listed showing the number of white residents residing within each state. The states represent nominal-level variables because there is no significant order to the list of states. In this case, they are listed alphabetically, but they could be listed in any other order without changing the significance of the report.
Ordinal-level:
An article on www.Forbes.com entitled “In Depth: America’s Most Popular Stores” lists several stores that have been rated by customers to indicate the level of customer satisfaction in several different categories. The rating scales for the categories are ordinal-level because there is a ranked order for the satisfaction scale (i.e., being extremely satisfied would be rated with a higher number than being extremely dissatisfied), but the magnitude of the difference between each rating cannot be determined.
Interval-level:
The statistical abstract for “Highest and Lowest Temperatures by State,” taken from the U.S. Census Bureau, lists the high and low temperatures for each state. The temperatures are interval-level variables because the values follow a ranked order (i.e., 80 degrees is lower than 90 degrees) and the differences between the values is constant (i.e., the difference between 80 degrees and 90 degrees is the same as the difference between 50 degrees and 60 degrees). However, the zero point does not represent absence of temperature; rather, it is merely a point on the scale.
Ratio-level:
The statistical abstract for “Federal Minimum Wage Rates” lists the federal minimum wage rates from 1938 through 2009. The wage rates are ratio-level variables because the values follow a ranked order (i.e., one dollar is greater than two dollars, etc.) and the difference between values is constant (i.e., the difference between 0.50 and 0.55 is the same as the difference between 1.25 and 1.30). In addition, the zero point would indicate that there was no federal minimum wage and significance can be drawn from the ratio between values (i.e., the rate in 1974 of $2.00 was twice as much as the rate of $1.00 in 1960).
Nominal-level:
Ordinal-level:
Interval-level:
Ratio-level: