ABSTRACT
Take a variable y which is taken to depend in a certain way on variables such as x1, x2, x3, etc., so that y=(x1, x2, x3, etc.). Experimenting on this system is the same as creating situations such as x1, x2, x3, etc., which have variable values from one situation to another. If the experimental programme (cf. Fisher) is well planned, it will allow us to define the influence of each of the ‘independent’ variables i.e., x1, x2, x3, etc., on the dependent variable y. Alternatively, the experiment may consist of the transformation of the structure f of the system linking y, x1, x2, x3, etc., in order to study the effect of a change from f to f’ on y.
