ABSTRACT
This chapter covers . . . . . . how to use dichotomies as independent variables in regression . . . how to use reference groups in regression . . . how researchers used reference groups to study gender differences in housework . . . how a researcher used reference groups to study attitude changes over time
INTRODUCTION
In the previous two chapters, all of the independent variables we have used have been ratio-level variables: height, income, hours, and grade point average, for example. Often we want to use other types of variables in regression that are not naturally measured at the ratio level. For example, we may want to use sex as an independent variable: do men and women signifi cantly differ on some dependent variable of interest? We (usually) consider sex to be a dichotomy: you are either male or female. Or we may want to compare people of various religious backgrounds: Catholic, Protestant, Jewish, Muslim, and so on. As you recall from Chapter 1 , religious background is a nominal-level variable: it simply puts people into categories without any means of ranking them (there’s no natural way to rank religious backgrounds). This chapter covers simple techniques we can use when we want to include these kinds of variables in a regression as independent variables.
