ABSTRACT
We turn now from the analysis of observations of a single variable to consider simultaneous observations on several variables. Instead of estimating differences caused by the application of different treatments, we will now focus on the relationship between two or more variables. There are various possible reasons for this focus, but the two principal ones are (1) to investigate and test hypothetical mathematical models for biological systems, and (2) to predict the values of one variable from another. Often a particular investigation may stem from both of these motives. Examples of possible investigations of multivariate relationships include:
• The response of a crop to varying amounts of several fertilizers. The objectives may be to establish the form of the relationship or to estimate the optimum combination of fertilizers.
