The Analysis of Discrete Data

Abstract

One consideration which has a very large influence on the methods of analysis which we adopt is the question of the type of measurement of which our observations are composed. Some possibilities are as follows:

• (i) Continuous data. This includes data such as the weight per unit length of plastic pipe, Example 1.3; measurements of height such as in Example 1.6, etc. In these cases it is convenient to think of the variables as being continuous, even though they must be recorded to a finite number of decimal places.

• (ii) Discrete data. This is when the observed values of our variables can only be one of a discrete set of values. Examples include the number of seeds germinating out of a box of 100, Example 1.1; the number of particles emitted from a radioactive source in a 5-second period, Example 1.2, etc.

• (iii) Ranked data. This is when, for example, a subject is asked to rank some drawings in order of preference; if the subject is asked to rank n items the observation on any item is one of the numbers, 1, 2, ..., n, representing the rank, and each of the numbers 1, 2,..., n appears just once in any ranking of n items. Methods of analysis for ranked data include those given in Chapter 7. (The methods of Chapter 7 can also be used on continuous data.)