Estimating the size of social networks

Abstract

It is notoriously difficult to define and estimate the size of individuals' social networks. Methods are suggested for estimating the size of such networks, using known rates of mobility together with survey data on the number of acquaintances of individuals who have recently moved. Special attention is given to the likely correlation between mobility rates and network size. Estimates are found to be consistent with previous estimates in the literature. The relation of this approach to multiplicity sampling and to "the small-world problem" is also discussed.