Mathematical models have become important to the routine functioning of commerce and government. Models help with planning, evaluating, managing and operating systems of all types. With the spread of personal computers, more models will be used by .. and developed by .. more people. Yet, even as society becomes more dependent on models, we have not given enough attention to the process of model formulation, either in practice or in education. Model formulation is the first step in the modeling process: moving from a vague sense of "problem" to a formal representation that can be manipulated to yield solutions. Improved, systematic methods of model formulation are important to the development of "modeling science" that will improve the efficiency and effectiveness of modeling efforts. The objectives of this research are to: (1) Describe the model formulation process as practiced by acknowledged experts; (2) Develop ways to judge the success of model formulation; (3) Suggest improved procedures for model formulation. This research will be exploratory, and should pave the way for experimentally verifiable propositions about effective model formulation. It should also lead to improved techniques and materials for educating modelers
