Modification of the Fisher's information measure to optimize a sampling design

In statistics, Fisher was the first to introduce the measure of the amount of information supplied by the data about the unknown parameter. We analyze the disadvantages of Fisher's information measure for optimization of sampling designs. To overcome this problem, we modify Fisher's information measure and we upgrade it to the multivariate setting. On a case study of soil we demonstrate the evaluation of different information measures derived from Fisher's information measure. The variables under study were the concentrations of several chemical compounds in soil (such as pH, N, C, Zn, etc.).