ABSTRACT
Previous chapters have presented the major theoretical formulations of coalition formation and payoff allocation in n-person cooperative games with sidepayments. Henceforth, interest focuses on the relationship between theoretical construction and experimental data. But which data? Creating an experimental game to study coalition formation behavior is no straightforward task, and virtually every investigator has his own idiosyncratic approach. The ensuing multitude of methodologies creates problems of interpretation and integration, because it is certain that the experimental formulation of the game affects the subjects’ interpretation of their task, their motivation, and their ideas of what the experimenter wishes of them, which consequently affect the negotiations and outcomes of the experimental games. Komorita and Meek (1978), for example, have shown that the same political convention game will support differing social psychological theories, depending on the experimental conditions of communication and information. The extent to which the results of an experiment may be used to explore the descriptive power of a theory is a function of the extent to which the assumptions of the theory are met by the experimental construction. Indeed, it has more than once been the case that an experimental paradigm is constructed with a specific theoretical orientation in mind (see, e.g., Gamson’s [1964] “advice to a coalition theorist”).
