2RV+HRV and Testing for Strong VS Full Dependence

TL;DR

This paper develops statistical tests to distinguish different types of dependence in network growth models, specifically focusing on identifying full, strong, and weak dependence based on limit measures.

Contribution

It introduces two new test statistics for dependence classification, analyzes their asymptotic behavior, and demonstrates their application with bootstrap methods on simulated and real data.

Findings

Test statistics effectively differentiate dependence types.

Asymptotic normality established for the test statistics.

Bootstrap methods validate the tests on practical data.

Abstract

Preferential attachment models of network growth are bivariate heavy tailed models for in- and out-degree with limit measures which either concentrate on a ray of positive slope from the origin or on all of the positive quadrant depending on whether the model includes reciprocity or not. Concentration on the ray is called full dependence. If there were a reliable way to distinguish full dependence from not-full, we would have guidance about which model to choose. This motivates investigating tests that distinguish between (i) full dependence; (ii) strong dependence (support of the limit measure is a proper subcone of the positive quadrant); (iii) weak dependence (limit measure concentrates on positive quadrant). We give two test statistics, analyze their asymptotically normal behavior under full and not-full dependence, and discuss applicability using bootstrap methods applied to simulated and real data.