隨機右設限存活資料之下根據聯合信賴域所做的多重比較; Multiple comparison procedures based on simultaneous confidence regions for random right-censored survival data

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/7476

Title:	隨機右設限存活資料之下根據聯合信賴域所做的多重比較;Multiple comparison procedures based on simultaneous confidence regions for random right-censored survival data
Authors:	王光祿;KuangLu Wang
Contributors:	統計研究所
Keywords:	多重比較;右設限;限制平均壽命;存活百分位數;分層比例風險模式;multiple comparison;right-censored;stratified proportional hazards model;survival percentile;restricted mean life time
Date:	2009-06-18
Issue Date:	2009-09-22 10:58:32 (UTC+8)
Publisher:	國立中央大學圖書館
Abstract:	多重比較法(multiple comparison procedure)是評估多個處理效應異同的多重決策。因聯合信賴域(simultaneous confidence region)除了能夠用作檢定，且可提供各處理組存活差異大小的訊息。所以，本文首先針對不具共變數的隨機右設限(right censored)存活資料，分別建構存活百分位數差異和限制平均壽命差異的聯合信賴域，然後，針對具共變數的隨機右設限存活資料，分別在分層Cox模式(stratified Cox model)和不分層Cox模式(unstratified Cox model)之下，建構條件存活百分位數差異和條件限制平均壽命差異的聯合信賴域。在比例風險和非比例風險二種模式，本文模擬研究上述所提信賴域的覆蓋機率(coverage)及平均距離(averaged absolute deviance)。最後，藉由實例分析說明本文所提各種方法的應用。 The multiple comparison procedures are multiple decisions that assess the differences of the effects for more than two treatment groups, simultaneously. Because the simultaneous confidence regions can not be only used for the problems of hypotheses testing, but also provide the information about the magnitudes of the differences between the survival indices of each treatment group. Therefore, for random right-censored survival data without covariates, the simultaneous confidence regions for the differences between the survival percentiles and the differences between the restricted mean life times are constructed, firstly. Next, for random right-censored survival data with covariates, the simultaneous confidence regions for the differences between the conditional survival percentiles and the differences between the conditional restricted mean life times are constructed under the stratified Cox model and the unstratified Cox model, respectively. In the simulation study, the cases of proportional hazards and nonproportional hazards are considered, and the coverages and the averaged absolute deviances between the lower bounds and the true parameters are then provided. Finally, the use of the proposed procedures in this article is illustrated with a right-censored survival data
Appears in Collections:	[Graduate Institute of Statistics] Electronic Thesis & Dissertation

Files in This Item:

File	Size	Format

社群 sharing

Loading...