On the aggregation across subjects in analyzing individual choice behavior

Authors dealing with models of individual choice behavior often point out the pitfalls of using group data to test such models. However, it is possible to derive properties of the choice behavior of the population based on these models of individual choice behavior. The present paper develops a procedure for the aggregation of choice data which takes into account individual differences. The utility of using aggregated choice data and the properties of the present procedure in comparison to other methods which utilized group data in analyzing choice behavior are discussed.