Valued Ties Tell Fewer Lies, II: Why Not To Dichotomize Network Edges With Bounded Outdegrees

TL;DR

This paper evaluates a method of binarizing valued networks by selecting top outbound ties and finds it performs worse than traditional thresholding, advising against its use.

Contribution

It introduces and empirically tests a new binarization method based on bounded outdegrees, demonstrating its limitations compared to standard thresholding.

Findings

Bounded outdegree method performs worse than thresholding.

Simulations and real data show the method increases network fragmentation.

Traditional thresholding remains preferable for valued network binarization.

Abstract

Various methods have been proposed for creating a binary version of a complex network with valued ties. Rather than the default method of choosing a single threshold value about which to dichotomize, we consider a method of choosing the highest k outbound arcs from each person and assigning a binary tie, as this has the advantage of minimizing the isolation of nodes that may otherwise be weakly connected. However, simulations and real data sets establish that this method is worse than the default thresholding method and should not be generally considered to deal with valued networks.