The Data Generating Model in Toxicity Test and Estimation for the Toxicity Indicators  

杀虫剂毒力测定中试验数据生成模型和毒力指标估计(英文)

作　　者：刘迎洲[1] 位刚[1] 卢恩双[1] 袁志发[1] 

机构地区：[1]西北农林科技大学理学院,陕西杨凌712100

出　　处：《Plant Diseases and Pests》2010年第4期42-45,共4页植物病虫害研究（英文版）

摘　　要：This paper gives the mathematical reason that the probit analysis method of toxicity measurement is reasonable, and proposes a new approach to compute the interval estimation of median lethal dose and 95% lethal dose. Based on the dose-response function of pesticides, this study firstly establishes a model of the data generating progress in toxicity test and proves that when the linear models of the logarithm of probability value and dose have been estimated, the weighted linear regression should be used, it is the reason why there is the heteroscedasticity of random disturbance term in the regression model, and the weight is right the reciprocal of the variance for random disturbance term. Secondly, based on the numerical simulation method, this paper gives a new approach for the interval estimation of median lethal dose and 95% lethal dose.This paper gives the mathematical reason that the probit analysis method of toxicity measurement is reasonable, and proposes a new approach to compute the interval estimation of median lethal dose and 95% lethal dose. Based on the dose-response function of pesticides, this study firstly establishes a model of the data generating progress in toxicity test and proves that when the linear models of the logarithm of probability value and dose have been estimated, the weighted linear regression should be used, it is the reason why there is the heteroscedasticity of random disturbance term in the regression model, and the weight is right the reciprocal of the variance for random disturbance term. Secondly, based on the numerical simulation method, this paper gives a new approach for the interval estimation of median lethal dose and 95% lethal dose.

关 键 词：Toxicity measurement Probit analysis BOOTSTRAP Estimation error Interval estimation 

分 类 号：S482.3[农业科学—农药学]