Statistics of the null distribution for a single node. Used for the normality approximation in p-value computation.