Stata

8 Logistic regression*

Linear regressions can be used, as long as the dependent variable is metric (examples of metric variables are wage, working hours per week or blood pressure). We often encounter data that does not fit this metric, as it is binary. Classical examples are: whether somebody is pregnant, voted in the last elections or has died, as only two possibilities (states) exist. This chapter will shortly introduce logistic regressions which are used when these types of variables should be explained. As we will not talk about any statistical foundations of the method, you are advised to have a look at the literature, for a general introduction (Acock, 2014: 329–345; Long and Freese, 2014). It is also recommended that beginners read chapter ...