Poisson及负二项分布对微核试验数据拟合效果  

Poisson distributon and negative binomial distribution fitting for data of cytokinesis-block micronucleus test

作　　者：郑辉烈[1,2] 王忠旭[3] 王增珍[1] 

机构地区：[1]华中科技大学同济医学院公共卫生学院流行病与卫生统计系,武汉430030 [2]南昌大学医学院公共卫生学院流行病与卫生统计学系 [3]中国疾病预防控制中心

出　　处：《中国公共卫生》2009年第9期1102-1104,共3页Chinese Journal of Public Health

基　　金：国家科技部"973"课题(2002CB512910)

摘　　要：目的探讨微核试验数据(每1000个双核淋巴细胞中具有微核的淋巴细胞数)的统计分布,为微核试验数据的统计分析提供依据。方法利用东北某大型钢铁公司158名焦炉工和66名非焦炉工微核检测的实例数据,进行矩法估计Poisson分布和二项分布的参数,同时拟合Poisson分布和负二项分布,比较其拟合效果。结果拟合优度检验显示,焦炉工和非焦炉工微核数据均不服从Poisson分布(P<0.001);焦炉工微核数据服从负二项分布(P=0.545),非焦炉工微核数据虽不服从负二项分布(P<0.001),但该数据似合负二项分布的效果要好于拟合Poisson分布(拟合负二项分布比值3.1<拟合Poisson分布比值20.3)。结论对于具有聚集性趋向(即方差与均数比值较大)的微核数据,相对于拟合Poisson分布,拟合负二项分布是一较好选择。Objective To investigate the statistical distribution of the data of cytokinesis-block micronucleus test to provide basis of statistical analysis. Methods The moment estimate parameters of the Poisson distribution, the negative binomial distribution, and Pearson goodness of fit test were used to evaluate the goodness of fit of the two different distributions. Results Neither the data of cytokinesis-block mucronucleus test for coke-oven workers and non-coke-oven workers presented Poisson distribution（ P 〈 0. 001 ）. The data of coke-oven workers showed a negative binomial distribution（ P = 0. 545） and the data of non-coke-oven workers did not（P 〈 0.001 ）. Conclusion For the data of cytokinesis-block micronucleus test with clustered characteristics , the negative binomial distribution fitting was a better choice compared to the Poisson distribution fitting.

关 键 词：微核 POISSON分布 负二项分布 似合优度 

分 类 号：R213[医药卫生—中医学]