Useful links
Unit of study_
BSTA5007: Linear Models
2021 unit information
The aim of this unit is to enable students to apply methods based on linear models to biostatistical data analysis, with proper attention to underlying assumptions and a major emphasis on the practical interpretation and communication of results. This unit will cover: the method of least squares; regression models and related statistical inference; flexible nonparametric regression; analysis of covariance to adjust for confounding; multiple regression with matrix algebra; model construction and interpretation (use of dummy variables, parametrisation, interaction and transformations); model checking and diagnostics; regression to the mean; handling of baseline values; the analysis of variance; variance components and random effects. NOTE: Linear Models is an important foundation unit. Students who do not develop a strong grasp of this material will struggle to become successful biostatisticians.
Unit details and rules
Managing faculty or University school:
Public Health
| Code | BSTA5007 |
|---|---|
| Academic unit | Public Health |
| Credit points | 6 |
|
Prerequisites:
?
|
BSTA5023 and (BSTA5011 or PUBH5010 or CEPI5100) |
|---|---|
|
Corequisites:
?
|
BSTA5002 |
| Prohibitions:
?
|
None |
| Assumed knowledge:
?
|
None |
At the completion of this unit, you should be able to:
- LO1. Have a sound understanding of the normal linear model including a theoretical grounding in the principles of least squares and likelihood-based estimation and related statistical inference, to the level of being able to manipulate equations required for deriving formulas for estimates and their standard errors for the standard models.
- LO2. Understand the principles and practice of model checking and diagnostics, and the use of transformations, in particular the log transformation, to improve model fit; understand the appropriate use of analysis of covariance to adjust for confounding; have a good working knowledge of the theory and practice of multiple regression analysis; be familiar with the method of analysis of variance (up to 2 factor models) and its relationship to multiple regression; gain an introductory understanding of nonparametric smoothing for flexible regression modelling, and of the use of variance components and random effects models.
- LO3. Have a strong grasp of practical issues involved in fitting linear models, including the ability to construct defensible models (use of dummy variables, choice of parameterisation, interaction and transformation of variables); demonstrate ability to fit models using modern statistical software and to interpret fitted models in terms that are useful to non-statisticians.
Unit availability
This section lists the session, attendance modes and locations the unit is available in. There is a unit outline for each of the unit availabilities, which gives you information about the unit including assessment details and a schedule of weekly activities.
The outline is published 2 weeks before the first day of teaching. You can look at previous outlines for a guide to the details of a unit.
There are no availabilities for this year.
Modes of attendance (MoA)
This refers to the Mode of attendance (MoA) for the unit as it appears when you’re selecting your units in Sydney Student. Find more information about modes of attendance on our website.
