| Credits |
3 credit points |
| Audience |
(Research) Master Human Movement Sciences |
| Lecturer |
Prof.dr. Aad van der Vaart
(Faculty of Sciences, VU). |
| E-mail: aad at few.vu.nl |
| Period |
April-May. First (Introduction) Meeting: Thursday April 5. |
| Hours 2012 |
Please check
http://www.rooster.vu.nl/ for the correct times. The following may be
outdated. |
|
Thursday 5 April: 15.30-17.15; Thursday 12, 19 April:
13.30-17.30; Monday 23 April: 9-13; Thursday 3, 10 May: 13.30-17.30;
Monday 14 May: 9-12.30; Tuesday 22 May: 13.30-15.15. |
|
MF-A311, except Thursdays 5 and 12 April MF-H625. |
| Aim |
An introduction to statistical models for regression analysis. |
| Form |
Lectures and computer assignments. |
| Description |
Regression models try to explain or predict a dependent variable
using measured "independent variables" (or "fixed
effects"). Statistical methods are needed if there is random variation
in the dependent variables and/or if the measured variables are a
sample from a population. Dependencies in the variables (for instance
by repeatingly measuring the same unit) require introduction of
unmeasured "random effects". In this course the concept of a
statistical model, as given by equations involving random variables,
will be central. The model reflects a design by which data is
collected, allows to formulate the assumptions underlying the
statistical analysis, and is basic for interpreting the results of the
analysis. The analysis itself is carried out by a computer package,
for which we need to know code but no formulas. |
| Brief introduction to basic statistical concepts.
Statistical model, likelihood function, maximum
likelihood estimation, confidence region, likelihood ratio test,
p-value. |
| Introduction to R. Basics of the open
source computer package R, and its application to fixed and mixed
effects regression. |
| Fixed effects regression.
Multiple linear regression: estimation, testing, variable
selection and diagnostics. Extension to generalized linear models. |
| Mixed effects regression.
Linear random effects models, repeated
measures, longitudinal data. Extension to nonlinear models. |
| Literature |
- Slides:
- Background text: Extending the Linear Model with R:
Generalized Linear, Mixed Effects and Nonparametric Regression Models,
by Julian J Faraway.
- R-manuals: a shorter one (in Dutch),
and the more extensive official one.
(See the R website for more.)
- Handouts
- Background reading for data analyses in lecture 5:
Appendix
of An R and S-PLUS Companion to Applied Regression by John Fox
(uses older version of lme4, with slightly different syntax).
- Other texts: Applied Longitudinal Analysis
by Fitsmaurice, Laird and Ware.
Mixed-effects models in S and S-plus by Pinheiro and Bates.
lme4: Mixed-effects Modeling with R by Bates (unfinished, draft chapters
can be downloaded via
lme website),
Linear Mixed Models for Longitudinal Data and
and Models for Discrete Longitudinal Data by Molenberghs and Verbeke.
|
| Assessment |
By weekly reports on data analyses using the R package. |
| Requirements |
Previous experience with statistics. Some knowledge of ANOVA.
Some knowledge of matrix algebra is handy. |
| Computer language |
The statistical package R can be downloaded from the R-project
site www.r-project.org .
It is free! |
| Assignments Spring 2012 |
- One , with supporting files
assign1.R,
LINREG1.R and HW1.rda.
- Two, with supporting file
assign2.R, and data file
fruitflies.txt.
- Three with supporting file
assign3.R and data file
and gala.txt.
- Four with supporting file
assign4.R, and
data files tlc1.txt,
tlc2.txt
and ashinalong.txt.
- Five, with supporting file
assign5.R, and
data file exercise2.txt.
- Six, with supporting file
assign6.R
- Seven, with supporting file
assign7.R.
|