Skriv ut | Lukk vindu |
Vår 2014
HEL-8002 Logistic regression and statistical analysis of survival data - 3 stp
The course is administrated by
Institutt for samfunnsmedisin
Course overlap
MED-8004 Logistic regression 1 stp
Course contents
Two main topics are covered:
- Simple, multiple and stepwise logistic regression, matched case-control studies, and ordinal logistic regression
- Methods for analysis of survival data. Includes the Kaplan-Meier survival estimator, the log rank test, and Cox's Proportional Hazard regression model.
Application deadline
20.01.2014
Admission requirements
Admission recommendations:
Introductory course in medical statistics. It is recommended that students who are planning to take HEL-8012 complete it before taking this course.
Objective of the course
- Analyze data using logistic regression models and models for the analysis of survival data (included Kaplan-Meier survival function, log rank test, and Cox proportional hazard regression).
- Identify different types of explanatory variables and correctly implement them in a logistic or Cox regression model.
- Interpret the results from logistic or Cox regression models.
- Assess interaction and confounding.
- Evaluate the model assumptions.
- Evaluate results from publications in medical journals where logistic or Cox regression models are applied, and critically assess the validity of its use.
Language of instruction
English
Teaching methods
The program consists of lectures, exercises with the use of PC and review of tasks. One can choose between the program packages SPSS, STATA and SAS.
Assessment
Exam and evaluation
Written assignment to be submitted. Evaluated with pass / fail.
Continuation exam
The course is held annually, no continuation exam.
Recommended reading/syllabus
The main literature is printed in a compendium. All content in the lectures and group exercises are part of the curriculum.
Current literature: David W. Hosmer, Applied Logistic Regression; David G. Kleinbaum, Survival Analysis and David G. Kleinbaum, Logistic Regression