Statistical Inference in the Case of Multiple Classification of Events

In the preceding section, we discussed the three types of statistical inference for the case of the simplest type of universe, namely, one in which the event in question could happen only in two alternative ways. We there arrived at the formulae for drawing conclusions from a comprehensive mass of data to partial masses drawn from it, for drawing conclusions in the opposite direction, that is, from a comparative small sample to the comprehensive mass of data from which the sample was taken, and finally for drawing conclusions from one sample to another, both of which were supposed to have been withdrawn from a comprehensive mass of data.