In earlier chapters, we described models for univariate data sets and developed confidence intervals and hypothesis tests for the parameters of these models.

In this chapter, we describe models for bivariate data sets in which a response depends on a numerical explanatory variable — the kind of data that we summarised earlier with a least square line. Linear regression models are closely related to least squares lines.

The parameters of these regression models are often of practical importance, and inference about their values helps us to interpret the relationship between the two variables.