This part deals with tests on proportions. After the Gaussian distribution the binomial distribution is probably the next most famous distribution. Binomial samples are very common and more intuitive than a Gaussian distribution. *Ill* and *healthy*, *success* and *failure*, *poor* and *rich* are well known binomial outcomes. The binomial distribution is linked to the normal distribution via large sample approximation and the normal distribution is the square root of the -distribution. More importantly, the binomial distribution is a special case of the multinomial distribution. This distribution plays a crucial role in the analysis of contingency tables and the tests in Chapter 4 can also be described in a contingency table set-up. However, this topic in general is covered in Chapter 14. Here we deal with well known special cases. Often the question occurs, if a proportion is the same as a predefined value or if two (or more) proportions differ significantly from each other.

Start Free Trial

No credit card required