Statistical Methods 2007-2008

This course aims to provide an overview of general statistical methods. It is joint course lectured by Prof. Brian Ripley, YY Teo and myself. It covers a wide range of material which falls into two general categories: 
 
        Basic methods of statistical analysis
        Experimental Design (or how data comes to be collected)  

As you would expect from a methodological course, it has a strong practical element to it. We try and motivate the course with examples arising from real life and the course is accompanied by weekly R practicals where we implement the techniques described in lectures. 
 
Lecture Notes on Statistical Analysis
 
Exploratory Data Analysis [PDF]
Hypothesis testing [PDF]
Robust Statistics [PDF]
Simulation  [PDF]
Linear models Part I 
Linear models Part II
Logistic and log-linear models
 

Lecture Notes on Experimental Design
 
Lecture notes on experimental design [PDF] 
Analysis of Sampling Plans
Some Designs
Screening Designs

 
R practicals
 
Some general advice on writing up your practicals.  

Week 1 - Introduction to using R
Week 2 - Hypothesis testing and some simple plots
Week 3 - Plots, test and simulation
Week 4 - Basic linear regression
Week 5 (Assessed) - Worksheet and dataset. Please note these will have restricted access until approximately 12 noon. 
Week 6 - Logistic and log-linear models with accompanying dataset Atomic. 
Week 7 - Logistic and log-linearpractical. 
 
 
Problem Sheets  

Exercises on lectures 1-4 and accompanying solutions
Exercises on Linear Models and accompanying solutions
Exercises on methods for count data and GLMs and accompanying solutions
 
Additional Notes
 
There is also some additional material on linear models
 
Formulae for linear models [PDF]   
Brief notes on contrasts
 
You may also wish to look at the Statistical Methods page from two years ago. Please be aware that much of the material on the linear models and experimental design aspects of the course is the same.

Dr Tim Heaton