Simulation models for psychometric theories

Abstract

My topic concerns another instance of a new methodology for the behavioral sciences made feasible by the availability of large fast computers. The strategy which characterizes this methodology is the attempt to simulate by means of computer programs, the paradigms which constrain and direct the behavior of an experienced behavioral scientist as he attempts to deal with the problems and data of interest to him. The significance for psychology of this computer-based methodology is similar to that of psychometric methodology to the extent that it provides formal systems for the explication and publication of the intuitions of the experienced psychologist.