Do Monte Carlo Simulations Help Us Better Understand and Explain Risk? D. Wartenberg, J. Bukowski, and M. A. Gallo, New Jersey Graduate Program in Public Health, UMDNJ-Robert Wood Johnson Medical School, Piscataway, NJ; and Environmental and Occupational Health Sciences Institute, Piscataway, NJ
Monte Carlo simulation methods are becoming an increasingly popular tool in quantitative risk assessment. They are used to quantify uncertainty in total risk estimates and as graphics for the communication of the precision of risk estimates to non-risk analysts. We are concerned that use of Monte Carlo methods may be misleading. They assess model precision but not model validity, they complicate use of sensitivity analysis, and they provide systematically lower risk estimates. While Monte Carlo methods may be useful in certain more limited applications, we argue against their widespread use in quantitative risk assessment.
This research was supported, in part, by a grant from the Hazardous Substances Management Research Center of the New Jersey Institute of Technology.