The null AZD-3463 that is to be tested, in this case, can be stated as: there is no evidence that one of the scenarios is intrinsically better than the other one, since the difference between their S-metric values could be explained as being a realization of an underlying probability distribution of the test object that is neutral (the distribution does not favor any scenario).
The central assumption of the method is that, if the observed result has arisen by chance, then this value will not seem unusual in a distribution of results obtained through many random relabelings of the samples. The randomization method proceeds as follows:
1.Compute the object (the mean difference between S-metric) for the samples for each scenario. This is the observed value;2.Reallocate randomly half of the samples to one scenario and half to the other. Compute the object value as before;3.Repeat the previous step until a large number of randomized object values have been generated and construct a distribution of these values;4.If the observed value is not more extreme than a fraction of the resulting outcomes ectoderm corresponds to a significance level, then consider the null hypothesis; otherwise reject it.