부산대학교 도서관

Cook and Stefanski have described a computer-intensive method,for approximately consistent estimation in regression problem with additive measurement error. In this article we derive the asymptotic distribution of their estimators and show how to compute estimated standard errors. These standard error estimators can either be used alone or as prepivoting devices in a bootstrap analysis. We also give theoretical justification to some of the phenomena observed by Cook and Stefanski in their simulations. KEY WORDS: Asymptotics, Bootstrap; Computationally intensive methods; Measurement error models.
The asymptotic distribution of the estimators derived by Cook and Stefanski's computer-intensive SIMEX method for approximately consistent estimation in regression problems with additive measurement error, is computed. The technique of calculating estimated standard errors is presented. These standard error estimators can either be utilized alone or as prepivoting devices in a bootstrap analysis.
l. INTRODUCTION We consider regression problems where some of the predictors are measured with additive error. The response is denoted by [upsilon] and the predictors by ([Zeta], [chi]), but [chi] [...]