The mode is the attribute of a variable that occurs most frequently. Technically, the mode is not calculated; it is observed. You find it by simply looking at the data and seeing which attribute of a variable occurs the most. In table 20.3b, we see that the modal value for education is 12 years (there are 8 out of 30 cases). This tells us that finishing high school is the most common level of education in our sample of 30 respondents.

All variables (nominal, ordinal, and interval) have modal values, but nominal variables can only have modal values. In table 20.3a, for example, we see that there are 18 men (GENDER = 1) and 12 women (GENDER = 2). The mode for gender, then, is male for this sample of 30 (box 20.1). (The mode, by the way, was female for the full survey of 609 respondents. When you work with small samples, fluctuations of this magnitude are normal.)

Many distributions have more than one mode, and bimodal distributions are quite common. In a rural community that has experienced a lot of out-migration, for example, the age structure is likely to be bimodal: There are young people hanging around who aren’t old enough to leave and old people who can’t find work in the city because of their age.