Towards a new way to measure power: Applying conjoint analysis to group decisions

This paper presents some preliminary development towards a methodology for measuring power in a group purchase decision. We define power as the capacity of an individual to change the probability of an alternative being chosen by expressing an opinion about that alternative. We present a linear model that relates the group members' preferences to the probability that a given alternative will be chosen. The model parameters reflect specific measures of power and can be estimated by conjoint analysis. We present the results of a pilot field study, conducted on a sample of purchasing managers, which favorably assesses certain psychometric properties of the resulting measures for dyads.