On alternative formulations of the small-world metric in complex networks

Small-world networks, i.e. networks displaying both a high clustering coefficient and a small characteristic path length, are obliquitous in nature. Since their identification, the "small-worldness" metric, as proposed by Humphries and Gurney, has frequently been used to detect such structural property in real-world complex networks, to a large extent in the study of brain dynamics. Here I discuss several of its drawbacks, including its lack of definition in disconnected networks and the impossibility of assessing a statistical significance; and present different alternative formulations to overcome these difficulties, validated through the phenospaces representing a set of 48 real networks.