Deconvolution Based Score Tests in Measurement Error Models
نویسنده
چکیده
Consider a generalized linear model with response Y and scalar predictor X. Instead of observing X, a surrogate W = X + Z is observed where Z represents measurement error and is independent of X and Y. The efficient score test for the absence of association depends on m(w) = E(XIW = w) which is generally unknown (Tosteson and Tsiatis, 1988). Assuming that the distribution of Z is known, asymptotically efficient tests are constructed using nonparametric estimators of m(w). Rates of convergence for the estimator of m(w) are established in the course of proving efficiency of the proposed test.
منابع مشابه
Detection of Outliers and Influential Observations in Linear Ridge Measurement Error Models with Stochastic Linear Restrictions
The aim of this paper is to propose some diagnostic methods in linear ridge measurement error models with stochastic linear restrictions using the corrected likelihood. Based on the bias-corrected estimation of model parameters, diagnostic measures are developed to identify outlying and influential observations. In addition, we derive the corrected score test statistic for outliers detection ba...
متن کاملLocal and omnibus goodness-of-fit tests in classical measurement error models.
We consider functional measurement error models, i.e. models where covariates are measured with error and yet no distributional assumptions are made about the mismeasured variable. We propose and study a score-type local test and an orthogonal series-based, omnibus goodness-of-fit test in this context, where no likelihood function is available or calculated-i.e. all the tests are proposed in th...
متن کاملTESTING FOR AUTOCORRELATION IN UNEQUALLY REPLICATED FUNCTIONAL MEASUREMENT ERROR MODELS
In the ordinary linear models, regressing the residuals against lagged values has been suggested as an approach to test the hypothesis of zero autocorrelation among residuals. In this paper we extend these results to the both equally and unequally replicated functionally measurement error models. We consider the equally and unequally replicated cases separately, because in the first case the re...
متن کاملRank tests and regression rank score tests in measurement error models
The rank and regression rank score tests of linear hypothesis in the linear regressionmodel are modified for measurement error models. The modified tests are still distribution free. Some tests of linear subhypotheses are invariant to the nuisance parameter, others are based on the aligned ranks using the R-estimators. The asymptotic relative efficiencies of tests with respect to tests in model...
متن کاملThe Invariance of Score Tests to Measurement Error
The linear measurement error model is an alternative to the classical regression model, in which we assume that the independent variables are subject to error. This assumption can cause statistical inferences and parameter estimators to di er dramatically from those obtained from the classical regression model. However, inferences may remain unchanged even though the independent variables are a...
متن کامل