Abstract This paper deals with comparing Pareto populations by proposing a one-way classification analysis in which effects are random. Explicit expressions for several characteristics of the Pareto response variable such as marginal distribution and hazard functions, mean, variance and intraclass correlation coefficient are derived. Estimations of the proposed model parameters and predictions with the minimum mean square error loss are also achieved. A procedure for testing homogeneity of populations is proposed. The performances of estimates and predictions are assessed numerically.

中文翻译：

---

用于比较帕累托种群的随机效应模型

摘要 本文通过提出一种效应是随机的单向分类分析来比较帕累托种群。导出了帕累托响应变量的几个特征的显式表达式，例如边际分布和风险函数、均值、方差和类内相关系数。还实现了对所提出的模型参数的估计和具有最小均方误差损失的预测。提出了一种测试群体同质性的程序。估计和预测的性能以数字方式评估。