Permutation-Based Type I Error Study of Distributed Lag Non-Linear Model for Associations between Temperature and Mortality
| Autor: | Chung-Li Wu, 吳崇立 |
|---|---|
| Rok vydání: | 2015 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 103 Objectives: In environmental epidemiology, distributed lag non-linear model (DLNM) has been widely adapted to examine if there is an association between the environmental factor and the health outcome. The feature of this model is that- it not only considers the lag effect between the environmental factor and the health outcome, but also takes into account the non-linear relationship. However, it is uncertain how many lag terms are needed to declare significance at the same time to conclude the lag effect. The aim of - this thesis is to realize how the family-wise error rate (FWER) varies in different settings of the parameters, and also compare the results with the adjusted p-value after Bonferroni’s correction. Methods: We implemented the permutation test of the distributed lag non-linear model to examine the association between temperature and mortality using the longitudinal data in Taipei, Taiwan. Results: Although the results vary with respect to different settings (lag and numbers of knots), our conclusions provide a simple guideline for concluding the overall significance according to the simulation results. Conclusions: Based on our results, the number of - lags needed to conclude the lag effect may varies in different settings of parameters. We also applied the DLNM to the longitudinal data in Taiwan and confirmed the association between temperature and mortality. Keywords: Distributed lag non-linear models, Family-wise error rate, Multiple testing, Permutation test |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|