Analysis of Current Status Data with Mismeasured Covariate

• Main Authors: Chieh-Ning Fu, 傅婕寧
• Other Authors: Chi-Chung Wen
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/08437706492063012560

碩士 === 淡江大學 === 數學學系碩士班 === 100 === While there is considerable work on covariate measurement error problem for right-censored survival data, regression analysis for current status data with mismeasured covariate has just started recently. In this thesis, we consider the conditional score approach of Wen and Chen (2012) for the analysis of errors-in-variable current status data. Distinct from Wen and Chen (2012), we employ a sieve method to approximate the baseline odds ratio function in the proportional odds model. Our proposal has advantages that the computation of conditional score estimates is more efficient and the variance estimation of regression parameter estimate is exempted from estimating the distribution of censoring time. The proposed method is evaluated through simulation studies and illustrated with a real data.