Approximate Power of Score Test for Variance Heterogeneity Under Local Alternatives in Nonlinear Models
Approximate Power of Score Test for Variance Heterogeneity Under Local Alternatives in Nonlinear Models
复制标题
DOI：
10.1016/j.csda.2005.06.013
复制
发表时间：
2003-10
期刊：
影响因子：
--
通讯作者：
Jinguan Lin;B. Wei
中科院分区：
文献类型：
--
作者：
Jinguan Lin;B. Wei
Lin和Wei [2003.非线性回归模型的异方差检验。通讯员Theory Methods 32，171-192]开发了非线性回归模型中异方差的得分检验，并通过Monte Carlo模拟研究了该检验的功效。本文的主要目的是提出一种方法来估计局部功率的得分测试，基于非中心χ 2近似的这种权力下连续的替代品。该方法也被推广到具有AR（1）误差的非线性模型。该方法适用于欧洲兔子数据中异方差性得分检验的局部幂计算问题[Ratkowsky，1983.非线性回归模型Marcel Dekker，纽约，pp. 108-110]。模拟研究表明，渐近逼近有限样本的情况下是好的，在很宽的范围内的参数配置。
Lin and Wei [2003. Testing for heteroscedasticity in nonlinear regression models. Comm. Statist. Theory Methods 32, 171–192] developed the score test for heteroscedasticity in nonlinear regression models and investigated the power of this test through Monte Carlo simulations. The main purpose of this paper is to present an approach for estimating the local power for the score test, based on a noncentral χ2approximation to such power under contiguous alternatives. The approach is also extended to nonlinear models with AR(1) errors. The methods are applied to the problem of local power calculations for the score tests of heteroscedasticity in European rabbit data [Ratkowsky, 1983. Nonlinear Regression Modelling. Marcel Dekker, New York, pp. 108–110]. Simulation studies are presented which indicate that the asymptotic approximation to the finite-sample situation is good over a wide range of parameter configurations.