Relative Frequency

Relative frequency refers to the percentage or proportion of times that a given value occurs within a set of numbers, such as in the data recorded for a variable in a survey data set. In the following example of a distribution of 10 values—1, 2, 2, 3, 5, 5, 7, 8, 8, 8—while the absolute frequency of the value, 8, is 3, the relative frequency is 30% as the value, 8, makes up 3 of the 10 values. In this example, if the source of the data has a wider range of possible scores than the observed values (such as a 0–10 survey scale), then it is permissible to report that some of possible values (e.g. 0, 4, 6, 9, and 10) were not observed in this set of data and that their respective relative frequencies were zero (i.e. 0%).

In survey research, the relative frequency is a much more meaningful number than is the absolute frequency. For example, in a news article using results from a poll of 800 citizens, it is more meaningful to know that approximately two thirds of them (67.5%) are dissatisfied with the job the president is doing than to know that 540 citizens who were polled think this way.

Relative frequency can be displayed in a frequency table—which displays each value in a distribution ordered from lowest to highest—along with the absolute and cumulative frequencies associated with each value. Relative frequency also can be displayed graphically in a bar graph (histogram) or pie chart.