THE NORM-INFORMATION-DISTANCE MODEL - A STOCHASTIC APPROACH TO PREFERENCE CHANGE IN GROUP-INTERACTION

The Norm-Information-Distance (NID) model is a formalized description of the process of change in collective judgment tasks. The model is derived in two steps. First, the probability that an individual within a group will change opinion is expressed as a function of relative subgroup size for the new choice, the informational attractiveness of that choice, and the number of alternatives intermediate between the original and the new opinion, which we label distance. Second, these individual transition probabilities are merged into a matrix of group constellation transition probabilities. The result is a stationary Markov process that predicts group constellation changes period-for-period during an experimental deliberation. The NID model is tested against data from Stasser and Davis (1981) to successfully show its ability to approximate the opinion change process in a two-response case. It is further tested against a multiple-response case using 225 male law and economics students at the University of Freiburg, Germany, Subjects were instructed to determine an appropriate sentence in a mock jury case. The NID model was applied to these data in several ways and was able to predict change in choice probabilities at any point during the process. Normative influence as measured by faction size appears to explain more of choice change than does informational attractiveness. Informational influence seems to work conditionally in the sense that it relates to only those alternatives that are presently held by at least one group member. The NID model has a high degree of flexibility and is expected to be useful in a variety of tasks and situations, in addition to those described in this article.