Influence Analysis on Crossover Design Model
碩士 === 國立中正大學 === 數學系統計科學研究所 === 100 === Crossover design model is a commom statistical model for the bioequivalence study. The evaluation of this study is usually based on the intra-subject variability. However, the result of this analysis can be affected by influential subjects or outliers. Chow a...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en_US |
| Published: |
2012
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/42178357344232502470 |
| Summary: | 碩士 === 國立中正大學 === 數學系統計科學研究所 === 100 === Crossover design model is a commom statistical model for the bioequivalence study. The evaluation of this study is usually based on the intra-subject variability. However, the result of this analysis can be affected by influential subjects or outliers. Chow and Tze (1990) discussed this issue based on the likelihood distance and estimates distance to detect the potential outliers for the bioequivalence study.
In this thesis, we develop the influence functions as an alternative approach for detection of influential subjects. Moreover, the comparison between the proposed approach and the methods proposed by Chow and Tze are investigated. One simulated data and two real data are provided to illustrate the results of these approaches.
|
|---|