Modeling performance of engineering controls when reductions are largest at the highest environmental concentrations of the hazardous contaminant.
Shulman-S; Mead-K; Mickelsen-L
Proceedings of the 2002 American Statistical Association, Section on Statistics and the Environment. Alexandria, VA: American Statistical Association, 2003 Mar; :3227-3232
Engineering controls are often compared outdoors using randomized pairs to determine if a new control design reduces worker exposure to airborne hazardous contaminants - when compared to an uncontrolled work environment. A common occurrence is that the ratio of means (of controlled-to-uncontrolled work environments) depends upon concentration. When the uncontrolled environment is at its highest concentrations, the engineering control has greatest impact and the ratio is smallest (reduction is largest). Such interaction may arise outdoors because of wind and other natural conditions that vary contaminant concentrations. The following approaches to model this interaction are compared: a.) model estimates reduction separately for upper 25% of uncontrolled samples and for lower 75%; b.) model -based on Tukey's one degree of freedom model for interaction; c.) model generalizes Tukey's model. The ability of the approaches to model the interaction is compared.
Engineering-controls; Statistical-analysis; Models; Computer-models; Analytical-models; Mathematical-models
4676 Columbia Parkway, MS- R3 , Cincinnati, Ohio, 45226, USA
Proceedings of the 2002 American Statistical Association, Section on Physical and Engineering Sciences