Bayesian Analysis of Perceived Social Networks
Johan Koskinen

Abstract
Measurement accuracy is an inherent problem in social network analysis. The issue of actor accuracy in the reporting of their interactions with others, was raised by Bernard, Killworth and Sailer (e.g. Bernard et al.,1980) and provoked extensive debate. Krackhardt (1987) later introduced the concept of Cognitive Social Structures and several methods for aggregating different actor reports on the network into a single graph, with the aid of which for example the congruence of reports could be gauged. A statistical model for aggregating separate reports into a single consensus network, with the additional benefit of allowing estimates of actor accuracy to be obtained in the process, was proposed by Batchelder, Kumbasar and Boyd (1997). The purpose here is to investigate this approach to the problem in a Bayesian framework. The Bayesian analysis yields posterior estimates of network characteristics and of perceiver ''accuracy'' as well as a measure of the degree of evidence in data for different models. The procedures are illustrated by an analysis of empirical data.

Keywords: Bayesian statistical modelling. Consensus analysis. Cognitive social structures (CSS). Measurement reliability. Social network analysis.