ABSTRACT
When binary data, or grouped binary data expressed as proportions, have a relatively simple structure, such as the data sets in Examples 1.1 and 1.2 of Chapter 1, questions of practical interest can usually be answered using standard statistical methods for estimating and testing hypotheses about a single proportion, and for comparing two or more proportions. Underlying these methods is the assumption that the observations have a binomial distribution, and so this distribution has a central role in the analysis of binary data. The binomial distribution was first derived by James Bernoulli, although it was not published until 1713, eight years after his death.
