Survival Analysis & Regression for Rates

Subject 505-972 (2009)

Note: This is an archived Handbook entry from 2009. Search for this in the current handbook

Credit Points:

12.50

Level:

9 (Graduate/Postgraduate)

Dates & Locations:

This subject has the following teaching availabilities in 2009:

September, - Taught on campus.

Pre-teaching Period Start	not applicable
Teaching Period	not applicable
Assessment Period End	not applicable
Last date to Self-Enrol	not applicable
Census Date	not applicable
Last date to Withdraw without fail	not applicable

Classroom

Timetable can be viewed here. For information about these dates, click here.

Time Commitment:

Contact Hours: One 4-hour lecture per week over the last 6 weeks of semester
Total Time Commitment: Students will be expected to undertake additional tasks, reading and preparation equivalent to an average of 80 to 90 hours of additional time commitment.

Prerequisites:

505-969 Epidemiology & Analytic Methods I or equivalent

505-970 Epidemiology & Analytic Methods II or equivalent

505-971 Linear and Logistic Regression

Corequisites:

See Prerequisites

Recommended Background Knowledge:

Non Allowed Subjects:

Core Participation Requirements:

Coordinator

Dr Lyle Gurrin

Contact

Centre for Molecular, Environmental, Genetic & Analytic Epidemiology

School of Population Health

Subject Overview:	This subject expands on Linear and Logistic Regression, introducing the use of rates and rate ratios and the analysis of censored time to event (survival) data. The focus is on methods for modelling the relationship between events measured over time, or censored time-to-event outcomes with a number of covariates, including Poisson regression and survival modelling using the proportional hazards model (Cox regression). Emphasis is on practical application and interpretation of results in the context of standard epidemiological study designs and particularly longitudinal studies. Further topics may include the use of flexible regression models to represent non-linear relationships. Practical work will use the statistical package Stata.
Objectives:	On completion of this subject, students are expected: To gain an understanding of generalized linear regression modeling of events over time and censored survival time data To gain familiarity with the topics of model building and prediction in the context of generalized linear models in epidemiology To develop a basic understanding of the role of regression modeling of rates and epidemiology, particularly in the context of longitudinal studies To learn practical skills in fitting and interpreting generalized linear regression models for count data over time (Poisson and Cox models) in the statistical computing package Stata To be introduced to the theory of generalized linear models
Assessment:	One 1,500 word written assignment on modelling rates using Poisson regression due mid-teaching period (30%); One 2,000 word written assignment on modelling time-to-event data using Cox regression due at the end of semester (40%); An end of semester examination (1.5 hour in length constituting 30% (1,500 words) of the total assessment) to be held in the University examination period.
Prescribed Texts:	BR Kirkwood & JAC Sterne, Essential Medical Statistics Second Edition, Blackwell Science, 2003. Special computer skills required: Students are expected to have experience using the Stata statistical package for multivariate analytic methods. Resources provided to distance students (applicable only to distance education subjects)
Recommended Texts:
Breadth Options:	This subject is not available as a breadth subject.
Fees Information:	Subject EFTSL, Level, Discipline & Census Date
Generic Skills:	-
Links to further information:	http://www.sph.unimelb.edu.au
Notes:	This subject is a group 1 elective in the Master of Public Health.
Related Course(s):	Master of Epidemiology Master of Public Health

Download PDF version.