OBJECTIVE: To explore the application of regression on order statistics (ROS) in estimating nondetects for food exposure assessment. METHODS: Regression on order statistics was adopted in analysis of cadmium residual data set from global food contaminant monitoring, the mean residual was estimated basing SAS programming and compared with the results from substitution methods. RESULTS: The results show that ROS method performs better obviously than substitution methods for being robust and convenient for posterior analysis. CONCLUSION: Regression on order statistics is worth to adopt,but more efforts should be make for details of application of this method.
OBJECTIVE: To explore the application of regression on order statistics (ROS) in estimating nondetects for food exposure assessment. METHODS: Regression on order statistics was adopted in analysis of cadmium residual data set from global food contaminant monitoring, the mean residual was estimated basing SAS programming and compared with the results from substitution methods. RESULTS: The results show that ROS method performs better obviously than substitution methods for being robust and convenient for posterior analysis. CONCLUSION: Regression on order statistics is worth to adopt,but more efforts should be make for details of application of this method.