．單位首頁
．系所簡介
．系主任
．強力師資
．研究所
．課程概況
．各類學程
．期刊論文
．招生資訊
．演講專區
．榮譽榜
．系學會
．系友天地
．就業訊息
97年2月22日(五) 3:00 ~4:00 P.M.
演講者姓名： 王清雲 教授
演講者服務單位： Division of Public Health Sciences, Fred Hutchinson Cancer Research Center
Poisson Regression with Measurement Error Using Instrumental Variables in Calibration Sample
Abstract
The Radiation Effects Research Foundation (RERF) provides very rich data on the atomicbomb survivors. Although some measurement error methods have been applied to adjusting for radiation measurement error for RERF data, further development of semiparametric or nonparametric methods is important in understanding the radiation effect to cancer or other outcome variables. Dosimetry data may be considered as a surrogate variable for the unobserved underlying radiation exposure. Biomarker data such as percentage of stable chromosome aberration can be treated as a type of surrogate variable for radiation. It may be treated as an instrumental variable for the unobserved radiation dosage. In this paper, we consider poisson regression when radiation covariates are measured with error. However, the magnitude of measurement error is unknown and no repeated surrogate measurements are available. A nonparametric estimator is proposed, and its finite sample performance is examined via simulation studies.
Keywords: biomarker data, calibration sample, corrected score, measurement error
