\( \newcommand{\bm}[1]{\boldsymbol{\mathbf{#1}}} \DeclareMathOperator{\tr}{tr} \DeclareMathOperator{\var}{var} \DeclareMathOperator{\cov}{cov} \DeclareMathOperator{\corr}{corr} \newcommand{\indep}{\perp\!\!\!\perp} \newcommand{\nindep}{\perp\!\!\!\perp\!\!\!\!\!\!/\;\;} \)

APTS Statistical Modelling Preliminary Material

Helen Ogden

based on material by Anthony Davison, Jon Forster, Dave Woods and Antony Overstall

Introduction

In order to get the most out of the APTS module on Statistical Modelling, students should have, at the start of the module, a sound knowledge of the principles of statistical inference and the theory of linear and generalised linear models. Students should also have some experience of practical statistical modelling in R.

The following reading and activities are recommended for students to (re)-familiarise themselves with these areas.

Statistical inference: It is recommended that students (re)-read the notes of the APTS module on Statistical Inference, available from the APTS website, and complete the assessment exercise (if they have not already done so). No further material is provided here.

Linear and generalised linear models: A student who has covered Chapters 8 and 10.1-10.4 of Statistical Models by A. C. Davison (Cambridge University Press, 2003) will be more than adequately prepared for the APTS module. For students without access to this book, the main theory is repeated below. The inference methodology described is largely based on classical statistical theory. Although prior experience of Bayesian statistical modelling would be helpful, it will not be assumed.

Preliminary exercises: Eight exercises are included throughout these notes.

Practical statistical modelling in R: Some practical exercises are also provided at the end of these notes to enable students to familiarise themselves with statistical modelling in R.