ABSTRACT
This chapter questions why we need logistic regression and also questions what logistic regression does. It argues how to interpret logistic regression results and explains what odds ratios are. The chapter shows that how researchers used logistic regression to study the policing of protest and suggests how researchers used logistic regression to study attitudes toward global warming. Score on an index of immigration support. Hours of television watched. Student grade point averages. They all have something in common: they are all ratio-level variables. They all start at zero and continue on from there: 1 hour, 2 hours, 5 hours, 10 hours, 24 hours of television a day. Much of the variation we want to explain involves ratio-level variables. In social research we often want to explain phenomena that involve not ratio-level variables but dichotomies. When we have a dichotomy as our dependent variable, we do not use regular linear regression. We use a special type of regression called logistic regression.
