應用多變量與長期資料分析
Applied Multivariate and Longitudinal Data Analysis
| 節 | 週二 |
|---|---|
7 15:30–16:20 | 應用多變量與長期資料分析 YS405 2 節連堂 |
8 16:30–17:20 |
* 根據陽明交大上課時間表所列
Course Description: This course focuses on applied multivariate extensions of generalized linear models (GLMs) with more complex data structure such as longitudinal (repeated) measures, matching, and hierarchical samples. Topics include design issues, random effect models, and generalized estimating equation (GEE) for GLMs with longitudinal data. Multivariate methods such as principal component analysis, factor analyses, and structure equation models will be introduced.
Prerequisite: Biostatistical Modeling and Data Analysis or consent of the instructor
Class materials and Computing: Class notes, references, and demonstration for data analysis using statistical packages can be downloaded from e3 platform. Students are free to use any statistical software package to complete the data analysis projects. Learning guide for using SPSS, SAS, and R will be provided on e3, including video for beginners of the above three packages and demonstration video for SPSS and example codes for SAS and R.
Class participation, Data analysis projects
| 週次 | 主題 |
|---|---|
| 第 1 週 | Course Introduction; Studies with correlated data and multivariate data |
| 第 2 週 | Introduction to the family of generalized linear models - Linear, Logistic, Poisson, negative binomial, and Gamma models - Interpretation of the model parameters with varying link functions - Cases studies |
| 第 3 週 | The Mid-Autumn Festival |
| 第 4 週 | Linear regression model with correlated measures – marginal models - Generalized estimating equation (GEE) (quasi-likelihood approach) |
| 第 5 週 | Linear regression model with correlated measures – random effect models - Linear mixed model and multi-level models (likelihood approach) (1) |
| 第 6 週 | Linear mixed model and multi-level models (likelihood approach) (2) |
| 第 7 週 | Data analysis session (1) |
| 第 8 週 | Logistic regression beyond binary outcome - Conditional logistic regression for matched data - Ordinal logistic regression - Multinomial logistic regression |
| 第 9 週 | Logistic regression with longitudinal data - Logistic regression using GEE - Generalized linear mixed model for longitudinal binary outcome |
| 第 10 週 | Poisson log-linear model with GEE |
| 第 11 週 | Data analysis session (2) |
| 第 12 週 | Principal component analysis |
| 第 13 週 | Factor Analysis |
| 第 14 週 | Structure equation models (SEM) |
| 第 15 週 | Data analysis session (3) |
| 第 16 週 | Final exam (take home) |
References: Diggle, Heagerty, Liang, & Zeger. Analysis of longitudinal data 2nd ed. 2002 Oxford science publications. Hosmer & Lemoshew “Applied logistic regression” 3rd ed. Liang and Zeger. Regression analysis for correlated data. Annual Review of Public Health 1993. 14:43-68. McCullagh and Nelder. Generalized linear models. 2nd ed, 1989, Chapman and Hall. Rencher & Christensen. Methods of multivariate analysis. 3rd ed. (2012) Kutner, Nachtsheim, Neter, and Li. Applied linear statistical models. 5th ed.