ABSTRACT
Throughout the last several chapters we’ve been examining the correlates of political knowledge. We discovered that not only can you predict political knowledge from a variety of demographics and communication variables but that most of these variables uniquely contribute to explaining individual differences in political knowledge. With the exception of gender, all the predictor variables we have considered thus far have been quantitative in that they quantify individual variations on some variable, such as how frequently a person discusses politics or how often a person’s watches the national network news. Even a dichotomous variable such as gender can be thought of as a crude measure of quantitative dimension, such as a person’s biological maleness (1 = lots, 0 = none). But what about a multicategorical variable-a categorical variable with several categories-such as a person’s political affiliation? Some people are Democrats, others are Republicans, and still others identify with no party or some other political group such as the Green party. There is no sensible way that people can be quantified on a variable such as political affiliation because, by its nature, political affiliation is not a quantitative dimension along which people vary. At yet we still may be interested in determining whether people who identify with different political parties vary in some way, such as how much they know about politics.
