Analysis of the angular distribution of cosmic objects

We investigate a method that reveals anisotropies in the angular distribution of cosmic objects. In particular, we investigate ensembles with similar to 1000 objects, which is the interesting case of some astronomical catalogs. Considering test ensembles generated with a variable degree of anisotropy, we calculate the probability of any pair of objects to be separated by a given angular distance and compare this result with the same probability for a purely isotropic ensemble. We show that the use of sub-ensembles of the original full sky test ensemble, namely partial catalogs containing objects in the polar cap regions, can reveal, at any scale, possible angular correlations in the original full sky distribution. We also show the robustness of this method by comparing it with the Kolmogorov-Smirnov and chi(2) statistical tests.