threshold at which to examine a network. For this reason, when investigating some property of interest, network studies should report results over a range of thresholds or edge densities to demonstrate this relationship, as well as the reliability of the findings. Typical threshold ranges begin at or above zero, to avoid negative and/or weak edges, and stop once the graph begins to fragment into components. When comparing the properties of two or more networks (e.g. networks from multiple developmental cohorts), it is useful to control for edge density, so that differences in network properties do not arise from differences in graph density. There is no agreed-upon method for comparing network properties, but common methods include simple non-statistical comparisons, t-tests of values at particular edge densities between cohorts, and comparison of parameters derived from growth curve fits of properties vs. threshold (Supekar et al., 2009).
