Bivariate categorical data sets are modelled using probability, so the chapter starts with further material about probability, including the concepts of marginal and conditional probability.
Contingency table summarise bivariate categorical data. In contingency tables, we are usually interested in whether the row and column classifications are 'independent'. The concept of independence is defined in terms of probabilities and a test is described for whether the classifications are indeed independent, based on sample data.