Mathematical Models in Education and Training

A nontechnical introduction to the state of the art in modeling education systems, prepared for RAND's Air Force technical training study. Input-output models are convenient for examining voluminous data on student flows for short periods or for relatively static institutions. Manpower planning models seem least useful. Optimization and simulation models are complementary: Optimization models show resource allocation choices explicitly, yield plans and priorities, and stimulate policy-level discussion. Simulations can help in management and short-term planning if enormous quantities of data are available. A model need not be realistic if it gives useful answers, nor is there one best way to model a situation. Simple models are best at our present level of understanding the educational process. Included are a review of the literature and a selective bibliography. Mathematical details are given in an appendix.