Markov chain analyses of communication processes in negotiation

Mara Olekalns, Melbourne Business School
Philip L. Smith, University of Melbourne
Laurie Weingart, Carnegie-Mellon University

Abstract

Markov chain analysis provides a way to investigate how the communication processes in dyadic negotiations are affected by features of the negotiating context and how, in turn, differences in communication processes among dyads affect the quality of the final settlement. In Markov models, the communication process is represented as a sequence of transitions between states, which describes how tactics are used and how they are reciprocated during the course of a negotiation. This article provides an introduction to Markov chain analysis and shows, using simulated data, how Markov chain models may be analyzed using widely-available loglinear modeling software. Model selection, assessment of the order of a chain, analysis of residuals, and sample size are discussed.