|
04341 Survey Sampling, Generalized Linear Models
| |
Danish title: Statistik 3
|
| Type: Å, Language: DDD |
| |
Credit points:
5 point |
|
| | | Previous course: C0405 |
| Offered by:
Department of Mathematical Modelling
(IMM) |
| No credit points with: C0405 |
| Prerequisite: 04040/C0410/04041/C0401 |
| Desirable: 04241/C0411 |
| Recommended semester:
4th -7th semester |
| Examination:
Written exam
(13 point scale
) |
| Remarks: Notes are obtained at IMM. |
| Contact person: |
Poul Thyregod, IMM, Building 321, Tel. +45 4525 3361 |
|
| Aim: To give the participants a broad introduction to the application of statistical methods with special emphasis on a number of specific methodological areas like planning of surveys and analysis of survey data, decomposition of hierarchical variation, analysis of multidimensional tables of counts, and analysis of failure-time and survival data. After completion of the course the students are expected to be able to design and analyze a sample survey of a simple, structured population, to assess whether the variation in given data may be decomposed into hierarchically organised components, to analyze data from simple dose-response experiments, to assess the structure in data concerning lifetime and analogous phenomena. |
| Contents: Models for sampling variation: Sampling strategies, including stratified sampling and cluster sampling, ratio- and regression estimates. Models for analysis of hierarchical variation: Analysis of components of variation for the normal distribution and for simple one-dimensional distributions. Decomposition of compound distributions. Bayes theorem. Prior and posterior distribution of parameters. Empirical Bayes methods. Preposterior analysis. Model for analysis of categorical data: Simple dose-response models, logistic regression. Analysis of one dimensional. responses under crossed and under hierarchical models. Analysis of multidimensional. responses, log-linear models. Models for analysis of lifetime data: Lifetime distributions. Censored observations, survival tables. Analysis of models with proportional hazard linear effects of covariates. |
|
|
