Possible typo in the textbook: Section 2.8.3

[mamcisaac]

The paragraph at the start of Section 2.8.3 appears to say "slope" where it should say "intercept":

Covariates xij in (2.3) can either be continuous or categorical variables. If a categorical independent variable has k categories then k−1 indicator variables should be used to index the categories provided the regression model contains a slope. If the regression model does not contain a slope, then exactly k indicator variables are required.

I believe this is meant to say that the model could be
$$y_i = \beta_0 + \sum_{k=1}^{K-1} \beta_k x_{ik} + \varepsilon_i$$ or $$y_i = \sum_{k=1}^{K} \beta_k x_{ik} + \varepsilon_i$$
depending on whether the intercept $\beta_0$ is included.